Introduction

A common use case in many business applications is to find if an element exists in any document or not. This article provide ways to find such documents and explain points that should be taken care of while designing a solution.

Solution

In general, existence of an element in a document can checked by using below XQuery.

cts:element-query(xs:QName('myElement'),cts:and-query(()))

Note the empty cts:and-query construct here. An empty cts:and-query is used to fetch all fragments.

Hence running below search query will bring back all the documents having element "myElement".

Wrapping the query in cts:not-query will bring back all the documents *not* having element "myElement" 

As a search using cts:not-query is only guaranteed to be accurate if the underlying query that is being negated is accurate from its index resolution, hence to check existence of a specific XPath, we need to index that XPath.
e.g. if you want to find documents having /path/1/A (and not /path/2/A) then you can create a field index for path /path/1/A and then use it in your query instead.

Things to remember

1.) Have unique element name in a single document i.e. try not to use same element name at multiple places within a document if they have different meaning for your use case. Either give them different element names or put them under different namespaces to remove any ambiguity. e.g. if you have element "table" at two places in a single document then you can put them both under different namespaces such as html:table & furniture:table or you can name them differently such as html_table & furniture_table.

2.) If element names are unique within a document then you don't need to create additional indexes. If element names are not unique within a document and you are interested in only a specific XPath then create path(field) indexes on those XPaths and use the same in your not-query.

Introduction

MarkLogic Server has shipped with full support for the W3C XML Schema specification and schema validation capabilities since version 4.1 (released in 2009).

These features allow for the validation of complete XML documents or elements within documents against an existing XML Schema (or group of Schemas), whose purpose is to define the structure, content, and typing of elements within XML documents.

You can read more about the concepts behind XML Schemas and MarkLogic's support for schema based validation in our documentation:

https://docs.marklogic.com/guide/admin/schemas

Caching XML Schema data

In order to ensure the best possible performance at scale, all user created XML Schemas are cached in memory on each individual node within the cluster using a portion of that node's Expanded Tree Cache.

Best practices when making changes to pre-existing XML Schemas: clearing the Expanded Tree Cache

In some cases, when you are redeploying a revised XML Schema to an existing schema database, MarkLogic can sometimes refer to an older, cached version of the schema data associated with a given document.

Therefore, it's important to note that whenever you plan to deploy a new or revised version of a Schema that you maintain, as a best practice, it may be necessary to clear the cache in order to ensure that you have evicted all cached data stored for older versions of your schemas.

If you don't clear the cache, you may sometimes get references to the old, cached schema references and as result, you may get errors like:

XDMP-LEXVAL (...) Invalid lexical value

You can clear all data stored in the Expanded Tree Cache in two ways:

1. By restarting MarkLogic service on every host in the cluster. This will automatically clear the cache, but it may not be practical on production clusters.
2. By issuing a call to xdmp:expanded-tree-cache-clear() command on each host in the cluster. You can run the function in query console or via REST endpoint and you will need a user with admin rights to actually clear the cache.

An example script has been provided that demonstrates the use of XQuery to execute the call to clear the Expanded Tree Cache against each host in the cluster:

Please contact MarkLogic Support if you encounter any issues with this process.

Related KB articles and links:

• Performance implications of updating Module and Schema databases
• Best Practices for schema definition
• MarkLogic Blog: Using XML Schema with MarkLogic Server
• MarkLogic Blog: Schema Agnosticism: What It Is and Why You Should Care
• Techniques for dividing tasks between hosts in a cluster

Summary

XDMP-ODBCRCVMSGTOOBIG can occur when a non-ODBC process attempts to connect to an ODBC application server.  A couple of reasons that this can happen is that there is an http application that has been accidentally configured to point to the ODBC port, or a load balancer is sending http health checks to an ODBC port. There are a number of common error messages that can indicate whether this is the case.

Identifying Errors and Causes

One method of determining the cause of an XDMP-ODBCRCVMSGTOOBIG error is to take the size value and convert it to Characters.  For example, given the following error message:

2019-01-01 01:01:25.014 Error: ODBCConnectionTask::run: XDMP-ODBCRCVMSGTOOBIG size=1195725856, conn=10.0.0.101:8110-10.0.0.103:54736

The size, 1195725856, can be converted to the hexadecimal value 47 45 54 20, which can be converted to the ASCII value "GET ".  So what we see is a GET request being run against the ODBC application server.

Common Errors and Values

Conclusion

XDMP-ODBCRCVMSGTOOBIG errors, do not affect the operation of MarkLogic Server, but can cause error logs to fill up with clutter.  Determining that the errors are caused by an http request to an ODBC port can help to identify the root cause, so the issue can be resolved.

Summary

Meters data can be a good resource for getting an approximation of the number of requests being managed by the server at a given time. It's also important to understand how Meters data is generated, should there be a discrepancy between the Meters samples, and the entries in the access log.

Meters Request Data

The Meters data is designed to record a sampling of activity, every few seconds. Meters data is not designed to accurately record server request rates much lower than every few seconds. Request rates are 15-second moving averages, recalculated every second and available in real time through the xdmp:host-status, xdmp:server-status and xdmp:forest-status built-in functions.

Meters Samples

The metering subsystem samples these real-time rates on the minute and saves the samples in the Meters database. Meters sampled data of events that occur less frequently than the moving average period will be lower than the number of access log entries. The difference between the two will depend on when the last event happened and when the sample was taken.

This mean that if an event happens once a minute, the request rate will rise when an event happens, but then decay away within a few seconds. If the sample is taken after the event has decayed, the saved meters data will be lower than the actual number of requests

Conclusion

The result of the Meters sampling method means that it is not unusual for Meters to under report the number of requests in certain circumstances.

Summary

In MarkLogic Server v7.0-2, the tokenizer keys, for languages where MarkLogic provides generic language support, were removed so that they now all use the same key. For example, Greek falls into this class of languages. This change was made as part of an optimization for languages in which MarkLogic Server has advanced stemming and tokenization support.  

Stemmed searches that include characters from languages that do not have advanced language support, performed on MarkLogic Server v7.0-2 or later releases, against content loaded on a version previous to v7.0-2, may not return the expected results.

Resolution

In order to successfully run these stemmed searches, you can either:

• Reindexing the database ; or
• Reinsert the affected documents (i.e. the documents that contain characters in languages for which MarkLogic Server only has generic language support).

If these are not possible in your environment, you can always run the query unstemmed.

An Example

The following example demonstrates the issue

1. On MarkLogic Server version 7.0-1, insert a document (test.xml) that contains the Greek character 'ε'.
2. Run this query 
   xdmp:estimate( cts:search( doc('test.xml'), 'ε')),
cts:contains( doc('test.xml'), 'ε')
3. The query will return the correct results: 1, true
4. Upgrade MarkLogic Server to version 7.0-3 or later and run the query again
5. The query will return incorrect results: 0, false 
6. Reindex the database and re-run the query
7. The query will return the correct result once again.
    

Introduction

HAProxy (http://www.haproxy.org/) is a free, fast and reliable solution offering high availability, load balancing and proxying for TCP and HTTP-based applications.

MarkLogic 8 (8.0-8 and above) and MarkLogic 9 (9.0-4 and above) include improvements to allow you to use HAProxy to connect to MarkLogic Server.

MarkLogic Server supports balancing application requests using both the HAProxy TCP and HTTP balancing modes depending on the transaction mode being used by the MarkLogic application as detailed below:

1. For single-statement auto-commit transactions running on MarkLogic version 8.0.7 and earlier or MarkLogic version 9.0.3 and earlier, only TCP mode balancing is supported. This is due to the fact that the SessionID cookie and transaction id (txid) are only generated as part of a multi-statement transaction.
2. For multi-statement transactions or for single-statement auto-commit transactions running on MarkLogic version 8.0.8 and later or MarkLogic version 9.0.4 and later both TCP and HTTP balancing modes can be configured.

The Understanding Transactions in MarkLogic Server and Single vs. Multi-statement Transactions in the MarkLogic documentation should be referenced to determine whether your application is using single or multi-statement transactions.

Note: Attempting to use HAProxy in HTTP mode with Single-statement transactions prior to MarkLogic versions 8.0.8 or 9.0.4 can lead to unpredictable results.

Example configurations

The following example configurations detail only the parameters relevant to enabling load balancing of a MarkLogic application, for details of all parameters that can be used please refer to the HAProxy documentation.

TCP mode balancing

The following configuration is an example of how to balance requests to a 3-node MarkLogic application using the "roundrobin" balance algorithm based on the source IP address. The health of each node is checked by a TCP probe to the application server every 1 second.

backend app
mode tcp
balance roundrobin
stick-table type ip size 200k expire 30m
stick on src
default-server inter 1s
server app1 ml-node-1:8012 check id 1
server app2 ml-node-2:8012 check id 2
server app3 ml-node-3:8012 check id 3

HTTP mode balancing

The following configuration is an example of how to balance requests to a 3-node MarkLogic application using the "roundrobin" balance algorithm based on the "SessionID" cookie inserted by the MarkLogic server.

The health of each node is checked by issuing an HTTP GET request to the MarkLogic health check port and checking for the "Healthy" response.

backend app
mode http
balance roundrobin
cookie SessionID prefix nocache
option httpchk GET / HTTP/1.1\r\nHost:\ monitoring\r\nConnection:\ close
http-check expect string Healthy
server app1 ml-node-1:8012 check port 7997 cookie app1
server app2 ml-node-2:8012 check port 7997 cookie app2
server app3 ml-node-3:8012 check port 7997 cookie app3

Summary

MarkLogic Server organizes Trusted Certificate Authorities (CA) by Organization Name.  Trusted Certificate Authorities are the issuers of digital certificates, which in turn are used to certify the public key on behalf of the named subject as given in the certificate.  These certificates are used in the authentication process by:

1. A MarkLogic Application Server configured to use SSL (HTTPS).
2. Any Web Client which is making a connection to a MarkLogic Application Server over HTTPS (in the case of SSL Client Authentication).

Example Scenarios

Consider the following example:

$openssl x509 -in CA.pem -text -noout
Certificate:
    Data:
        Version: 3 (0x2)
        Serial Number: 18345409437988140316 (0xfe97fcaf8a61b51c)
    Signature Algorithm: sha1WithRSAEncryption
        Issuer: C=US, ST=CA, L=San Carlos, O=MarkLogic Corporation, OU=Engineering, CN=MarkLogic CA
        Validity
            Not Before: Nov 30 04:08:31 2015 GMT
            Not After : Nov 29 04:08:31 2020 GMT
        Subject: C=US, ST=CA, L=San Carlos, O=MarkLogic Corporation, OU=Engineering, CN=MarkLogic CA

In this example, From viewing the Trusted CA Subject field, the CA Certificate name will be listed with the organisation name of "MarkLogic Corporation" (O=MarkLogic Corporation) in MarkLogic's list of Certificate Authorities.

You can view the full list of currently configured Trusted Certificate Authorities by logging into the MarkLogic administration Application Server (on port 8001) and viewing the status page: Configure -> Security -> Certificate Authorities

Trusted CA Certificate without Organization name (O=)

In some cases, there are legitimate Trusted CA Certificates which do not contain any further information about the Organization responsible for the certificate.

The example below shows a sample self signed root CA (DemoLab CA) which highlights this scenario:

$openssl x509 -in DemoLabCA.pem  -text -noout
Certificate:
    Data:
        Version: 3 (0x2)
        Serial Number: 12836463831212471403 (0xb22447d80f91b46b)
    Signature Algorithm: sha1WithRSAEncryption
        Issuer: CN=DemoLab CA
        Validity
            Not Before: Nov 30 05:23:13 2015 GMT
            Not After : Nov 29 05:23:13 2020 GMT
        Subject: CN=DemoLab CA

If this Certificate were to be loaded into the MarkLogic, no name would appear under the list of <em>Certificate Authorities</em>in the list provided through the administration Application Server at Configure -> Security -> Certificate Authorities

In the case of the above example, it would be difficult to use the certificate validated by DemoLab CA (and to use DemoLab CA as our Trusted Certificate Authority) as MarkLogic will only list certificates that are associated with an Organization.

Solution

To workaround this issue, we can configure MarkLogic to use the certificate through some scripting with Query Console.

1) Loading the CA using Query Console

Start by using a call to pki:insert-trusted-certificates to load the Trusted CA into MarkLogic.  The sample Query Console code below demonstrates this process (Please ensure this query is executed against the Security database)

Make a note of value of the id returned by MarkLogic. It will return an unsigned long (xs:unsignedLong) which is the id value that can be used later to retrieve that certificate

2) Attach Trusted CA with "SSL Client Certificate Authorities" using Query Console

The next step is to associate the certificate that we just inserted from our filesystem (DemoLabCA.pem) with a given MarkLogic Application Server. Once this is done, any client connecting to that application server over SSL will be presented with the cerificate and DemoLab CA will be used to match the certificate using the Common Name value (Common Name eq "DemoLab CA")

3) Verify attached Trusted CA for Client Cetificate Authorities

Executing the above code should return the same identifier (for the Trusted CA) as returned as result of the code executed in step 1. Additionally, we can see that our Application Server (DemoAppServer) is now configured to expect an SSL Client Certificate Authority signed by DemoLab CA.

Further Reading

• Definition of a CA (on Wikipedia)
• Definition of a digital certificate (on Wikipedia)
• MarkLogic API documentation for pki:insert-trusted-certificates
• MarkLogic API documentation for admin:appserver-get-ssl-client-certificate-authorities
• MarkLogic API documentation for admin:appserver-set-ssl-client-certificate-authorities

Introduction

MarkLogic Server is engineered to scale out horizontally by easily adding forests and nodes. Be aware, however, that when adding resources horizontally, you may also be introducing additional demand on the underlying resources.

Details

On a single node, you will see some performance improvement in adding additional forests, due to increased parallelization. This is a point of diminishing returns, though, where the number of forests can overwhelm the available resources such as CPU, RAM, or I/O bandwidth. Internal MarkLogic research (as of April 2014) shows the sweet spot to be around six forests per host (assuming modern hardware). Note that there is a hard limit of 1024 primary forests per database, and it is a general recommendation that the total number of forests should not grow beyond 1024 per cluster.

At cluster level, you should see performance improvements in adding additional hosts, but attention should be paid to any potentially shared resources. For example, since resources such as CPU, RAM, and I/O bandwidth would now be split across multiple nodes, overall performance is likely to decrease if additional nodes are provisioned virtually on a single underlying server. Similarly, when adding additional nodes to the same underlying SAN storage, you'll want to pay careful attention to making sure there's enough I/O bandwidth to accommodate the number of nodes you want to connect.

More generally, additional capacity above a bottleneck generally exacerbates performance issues. If you find your performance has actually decreased after horizontally scaling out some part of your stack, it is likely that a part of your infrastructure below the part at which you made changes is being overwhelmed by the additional demand introduced by the added capacity.

Summary

MarkLogic Application Servers will keep a connection open after completing and responding to a request, waiting for another new request, until the Keep Alive timeout expires. However, there is an exception scenario where the connection will close regardless of timeout settings when the content is larger then 1 MB. This article is intended to provide further insight into connection close with respect to Payload size.

HTTP Header

Connection-Length

In general, Application Servers communicating in HTTP send the Content-Length header as part of their response HTTP Headers to indicate how many bytes of data the client application should expect to receive. For example

HTTP/1.1 200 OK

Content-type: application/sparql-results+json; charset=UTF-8

Server: MarkLogic

Content-Length: 1264

Connection: Keep-Alive

Keep-Alive: timeout=5

This requires Application Servers to know the length of the entire response data before the very first bytes (Response HTTP Headers) are put on to the wire. For small amounts of data, the time to calculate the content-length is fast; For large amounts of content, the calculation may be time consuming with the extreme being that the client finds the server unresponsive due to the delay in calculating the entire response length. Additionally, the server may need to bring the entire content into Memory Buffer, putting further burden on server resources.

Chunked-encoding

To allow servers to begin transmitting dynamically-generated content before knowing the total size of that content, HTTP 1.1 supports chunked encoding. This technique is widely used in music & video streaming and other industries. Chunked encoding eliminates the need of knowing the entire content length before sending a portion of the data, thus making the server looks more responsive.

At the time of this writing, MarkLogic Server (v8.0-6 and earlier releases) does not support chunked encoding. However, do look for this feature in future releases of MarkLogic Server.

Connection Close

In MarkLogic Server v7 and v8, MarkLogic Server closes the connection after transmitting content greater 1MB, which allows MarkLogic to avoid calculating content length in advance. The client will not see Content-Length Header for Larger (>1MB) content in HTTP Response from MarkLogic. Instead it will receive a Connection Close header in HTTP Response. After sending the entire content, MarkLogic Server will terminate the connection, to indicate to Client that the end of content has been reached.

Closing the existing connection for content larger then 1MB is an exception to the Keep-Alive configuration. This may result in unexpected behavior on clients that relying on MarkLogic Server respecting the Keep-Alive configuration, so this behavior should be accounted while designing Client Application Connection Pool.

Client Applications may have to send TCP SYN again to establish new connection to send subsequent request, which will add overhead of TCP 3 way handshake before sending next request. However, in the context of the data transfer for larger payload (>1MB), where many more round trips are added in overall communication, overhead of TCP 3 way handshake is very nominal.

Further Reading

• Chunked transfer encoding (Wikipedia)
• KB - "Notes on Using the Java Client API efficiently"

Summary

CSV files are a very common data exchange format. It is often used as an export format for spreadsheets, databases or any other application. Depending on the application, you might be able to change the delimiter character to a #hash or *asterix etc. One of the default delimiter definitions is a tab character. Content Pump supports reading and loading such CSV files.

Detail

The Content Pump -delimiter option defines which delimiter will be used to split the columns. Defining a tab as a value for the delimiter option on the command line isn't straight forward.

Loading tab delimited data files with content pump can result in an error massage like the following:

mlcp>bin/mlcp.sh IMPORT -host localhost -port 9000 -username admin -password secret -input_file_path sample.csv -input_file_type delimited_text -delimiter '    ' -mode local
13/08/21 15:10:20 ERROR contentpump.ContentPump: Error parsing command arguments: 
13/08/21 15:10:20 ERROR contentpump.ContentPump: Missing argument for option: delimiter
usage: IMPORT [-aggregate_record_element <QName>]
... 

Depending on the command line shell, a tab needs to be escaped to be understand from the shell script: 

On bash shell, this should work: -delimiter $'\t'
On Bourne shell, this should work: -delimiter 'Ctrl+V followed by tab' 
Alternative way would be to use: -delimiter \x09 

If none of these work, another approach you can try is to use the -options_file /path/to/options-file parameter. The options file can contains all of the same parameters as the command line does. The benefit of using an option file is that the command line is simpler and characters are interpreted as intended. The options file will contain multiple lines where the first line is always the action like IMPORT,  EXPORT etc. followed by a pair of lines. The first line is the option parameter and second the value for the option.

A sample could look like the following:

Make sure the file is saved in UTF-8 format to avoid any parsing problems. To define a tab as delimiter, place a real tab between single quotes (i.e. '<tab>')

To use this option file with mlcp execute the following command:

Linux, Mac, Solaris:

mlcp>bin/mlcp.sh -options_file /path/to/sample.options

Windows:

mlcp>bin/mlcp.bat -options_file /path/to/sample.options

The options file can take any paramter which mlcp understands. It is important that the action command is defined on the first line. It is also possible to use both command line parameters and the option file. Command line parameters take precedence over those defined in the options file.

Summary

There are sometimes circumstances where the MarkLogic data directory owner can be changed.  This can create problems where MarkLogic Server is unable to read and/or write its own files but is easily corrected.

MarkLogic Server user

There are sometimes circumstances where the MarkLogic data directory owner can be changed; this can create problems where MarkLogic Server is unable to read and/or write its own files.

The default location for the data directory on Linux is /var/opt/MarkLogic and the default owner is daemon.

If you are using a nondefault (non-daemon) user to run MarkLogic, for example mlogic, you would usually have 

    export MARKLOGIC_USER=mlogic

in 

    /etc/marklogic.conf 

Correct the data directory ownership

If the file ownership is incorrect, the way forward is to change the ownership back to the correct user.  For example, if using the default user daemon:

1.  Stop MarkLogic Server.

2.  Make sure that the user you are using is correct and available on this machine.

3.  Change the ownership of all the MarkLogic files (by default /var/opt/MarkLogic and any/all forests for this node) to daemon.  The change needs to be made recursively below the directory to include all files.  Assuming all nodes in the cluster run as daemon, you can use another unaffected node as a check.  You may need to use root/sudo permissions to change owner.  For example:

chown -R daemon:daemon /var/opt/MarkLogic

4.  Start MarkLogic Server.  It should now come up as the correct user and able to manage its files.

References

• MarkLogic Documentation: Installing MarkLogic

Introduction:

MarkLogic Server allows you to set-up an alerting application to notify users when new content is available that matches a predefined query. This can be achieved through the Alerting API with the Content Processing Framework (CPF). CPF is designed to keep state for documents, so it is easy to use CPF to keep track of when a document in a particular scope is created or updated, and then perform some action on that document. However, although alerting works for document updates and inserting, it does not occur for document deletes. You will have to create a custom CPF pipeline to catch the delete through an appropriate status transition.

Details

To achieve alerting for document delete, you will have to write your own custom pipeline with status transition to handle deletes. For example:

The higher 'priority' value and 'always' = true indicates that the custom pipeline has precedence over the default status change handling pipeline to handle document deletes.  Similarly, in the action module, you can write your custom code for alerting.

Note: By default, when a document is deleted, the on-delete pre-commit trigger is fired and it calls the action in the Status Change Handling pipeline (if enabled) for ‘delete’ status transition. It is recommended that you do not modify this pipeline as it can cause compatibility problems in future upgrades and releases of MarkLogic server.

Introduction

If you're looking at the MarkLogic Admin UI on port 8001, you may have noticed that the status page for a given database displays the last backup dateTime for a given database.

We have been asked in the past how this gets computed so the same check can be performed using your own code.

This Knowledgebase article will show examples that utilise XQuery to get this information and will explore the possibility of retrieving this using the MarkLogic ReST API

XQuery: How does the code work?

The simple answer is in the forest status for each of the forests in the database (note these values only appear if you have created a backup already).  For the sake of these examples, let's say we have a database (called "test") which contains 12 forests (test-1 to test-12).  We can get the backup status for these using a call to our ReST API:

http://localhost:8002/manage/LATEST/forests/test-1?view=status&format=html

In the results returned, you should see something like this:

last-backup : 2016-02-12T12:30:39.916Z datetime
last-incr-backup : 2016-02-12T12:37:29.085Z datetime

In generating that status page, what the MarkLogic code does is to create an aggregate: a database doesn't contain documents in MarkLogic; it contains forests and those forests contain documents.

Continuing the example above (with a database called "test" containing 12 forests) if I run the following:

This will return the forest status(es) for all forests in the database "test" and return the forest names using XPath, so in this case, we would see:

<forest-name xmlns="http://marklogic.com/xdmp/status/forest">test-1</forest-name>
[...]
<forest-name xmlns="http://marklogic.com/xdmp/status/forest">test-12</forest-name>

Our admin UI is interrogating each forest in turn for that database and finding out the metrics for the last backup.  So to put that into context, if we ran the following:

This gives us:

<last-backup xmlns="http://marklogic.com/xdmp/status/forest">2016-02-12T12:30:39.946Z</last-backup>
[...]
<last-backup xmlns="http://marklogic.com/xdmp/status/forest">2016-02-12T12:30:39.925Z</last-backup>

The code (or the status report) doesn't want values for all 12 forests, it just wants the time the last forest completed the backup (because that's the real time the backup completed), so our code is running a call to fn:max:

Which gives us the max value (as these are all xs:dateTimes, it's finding the most recent date), which in the case of this example is:

2016-02-12T12:30:39.993Z

The same is true for the last incremental backup (note all that we're changing here is the XPath to get to the correct element:

So we can get the max value for this by getting the most recent time across all forests:

This would give us 2016-02-12T12:37:29.161Z

Using the ReST API

The ReST API also allows you to get this information but you'd need to jump through a few hoops to get to it; the ReST API status for a given database would give you the names of all the forests attached to that database:

http://localhost:8002/manage/LATEST/databases/test

And from there you could GET the information for all of those forests:

http://localhost:8002/manage/LATEST/forests/test-1?view=status&format=html
[...]
http://localhost:8002/manage/LATEST/forests/test-12?view=status&format=html

Once you'd got all those values, you could do what MarkLogic's admin code does and get the max values for them - although at this stage, it might make more sense to write a custom endpoint that returns this information, something like:

Where you could make a call to that module to get the aggregates (e.g.):

http://[server]:[port]/[modulename.xqy]?db=test

This would return the database status for any given parameter-name that is passed in.

Problem:

When searching for matches using OR'ed word-queries, and in the case where there are overlapping matches, (i.e. one query contains the text of another query), the results of a cts:highlight query are not as desired.

For example:

 The desired outcome of this would be:

               <p>From the <m>memoirs of an accomplished artist</m> </p>

 Whereas, the actual results are:

                <p>From the <m>memoirs of an </m> <m>accomplished artist</m></p>

This behavior is by design and the results are expected. It is because cts:highlight  breaks up overlapping  areas into separate matches.

The cts:highlight built-in variables – $cts:queries and $cts:action help in understanding how this works, as well as to work-around this problem.

  $cts:queries --> returns the matching queries for each of the matched texts.

  $cts:action --> can be used with xdmp:set to specify what should happen next

• "continue" - (default) Walk the next match. If there are no more matches, return all evaluation results.
• "skip" - Skip walking any more matches and return all evaluation results
• "break" - Stop walking matches and return all evaluation results

   For eg., replacing the return statement with the following in the original query:

These results give us a better understanding of how the text is being matched. We can see that " accomplished artist" is matched by both the word-queries 'accomplished artist' and 'memoirs of an accomplished artist'; hence the results of cts:highlight seem different.

To work around this problem, we can insert a small piece of code: 

Please note that this solution relies on assumptions about what's inside the or-query, but this example could be modified to handle other overlapping situations.

Summary

Long URI prefix may lead to imbalance in data distribution among the forests. 

Observation

Database assignment policy is set to 'Bucket'. Rebalancer is set to enable, and no fragments is pending to be rebalanced; However, data is imbalanced across forests associated with database. Few forests has higher number of fragments compared to other forests in a given database.

Root cause

For bucket assignment policy, document uri is hashed to match specific bucket. The bucket policy algorithm maps a document’s URI to one of 16K “buckets,” with each bucket being associated with a forest. A table mapping buckets to forests is stored in memory for fast assignment.

Bucket algorithm does not consider whole uri length for the calculation while determining bucket based on uri hash. Uri based bucket determination in bucket assignment policy rely largely on initial characters for hashing algorithm.

If document uri includes long common prefix then all documents uri will result in same hash value and same bucket, even if they different suffix number, and hence result is skewed if there is larger common prefix.

Analysis

To confirm if uneven number of fragments between different forests in database, you can run below query which will give 100 sample documents from each forests and you can review if there are common prefix in document uri in forests with higher number of fragments.

xquery version "1.0-ml";

for $i in xdmp:database-forests(xdmp:database('<dbname>'))
    let $uri := for $j in cts:uris((),(),(),(), $i)[0 to 100]
                return <uri>{$j}</uri>
return <forests><forest>{$i}</forest><uris>{$uri}</uris></forests>

Recommendation

We recommend document uri to not have long name and common prefix. Certain common document uri values can be changed to collection.

Example uri -  /Prime/InternationalTradeDay/Activity/AccountId/ABC0001/BusinessDate/2021-06-14/CurrencyCode/USD/ID/ABC0001-XYZ-123.json

Can be -  /ABC0001-XYZ-123.json. with collection "USD", "Prime", and doc that have date element with "2021-06-14".

Above is just an example, but suggestion is to have an URI naming pattern to avoid large common prefix or save under collection. 

You can use document-assign built-in to verify if URI’s are distributed per the bucket algorithm.

https://docs.marklogic.com/xdmp:document-assign

Additional Resources

• Bucket Assignment Policy
• Databse Rebalancing
• Data Ingestion
• Organizing Documents

Introduction

In the Scalability, Availabilty & Failover Guide, the node communication section describes a quorum as >50% of the nodes in a cluster.

Is it possible for a database to be available for reads and writes, even if a quorum of nodes is not available in the cluster?

The answer is yes, there are configurations and sequences of events that can lead to forests remaining online when there are fewer than 50% of the hosts being online.

Details

If a single forest in a database is not available, the database is not be accessible. It is also true that as long as all of a database's forests are available in the cluster, the database will be available for reads and writes regardless of any quorum issues.

Of course, the Security database must also be available in the cluster for the cluster to function.

Forest Availability: Simple Case

In the simplest case, if you have a forest that is not configured with either local disk failover or shared disk failover and as long as the forest's host is online and exists in the cluster, the forest will be available regardless of any quorum issues.

To explain this case in more detail: if we have a 3-node MarkLogic cluster containing 3 hosts (let's call them host-a, host-b and host-c); if we were to then initialize host-a as the primary host (so this is the first host is set up in the cluster and is the host containing the master security database) and we then join host-b and host-c to host-a to complete the cluster. 

Shortly after that, if we shut both the joiner hosts (host-b and host-c) down, so only host host-a remained online, we would see a chain of messages in the primary host's ErrorLog that indicated there was no longer quorum within the cluster:

2020-05-21 01:19:14.632 Info: Detected quorum (3 online, 1 suspect, 0 offline)
2020-05-21 01:19:18.570 Warning: Detected suspect quorum (3 online, 2 suspect, 0 offline)
2020-05-21 01:19:29.715 Info: Disconnecting from domestic host host-b.example.marklogic.com because it has not responded for 30 seconds.
2020-05-21 01:19:29.715 Info: Disconnected from domestic host host-b.example.marklogic.com
2020-05-21 01:19:29.715 Info: Detected suspect quorum (2 online, 1 suspect, 1 offline)
2020-05-21 01:19:33.668 Info: Disconnecting from domestic host host-c.example.marklogic.com because it has not responded for 30 seconds.
2020-05-21 01:19:33.668 Info: Disconnected from domestic host host-c.example.marklogic.com
2020-05-21 01:19:33.668 Warning: Detected no quorum (1 online, 0 suspect, 2 offline)

Under these circumstances, we would be able to access the host's admin GUI on port 8001 and it would respond without issue.  We would be able to access Query Console on that host on port 8000 and would be able to inspect the primary host's databases.  We would also be able to access the Monitoring History on port 8002 - all directly from the primary host.

In this scenario, because the primary host remains online and the joining hosts are offline; and because we have not yet set up failover anywhere, there is no requirement for quorum, so host-a remains accessible.

If host-a also happened to have a database with forests that only resided on that host, these would be available for queries at this time.  However, this is a fairly limited use case because in general, if you have a 3-node cluster, you would have a database whose forests reside on all three hosts in the cluster with failover forests configured on alternating hosts. 

As soon as you do this, if you lose one host and you don't have failover configured, the database would now become unavailable (due to a crucial forest being offline) and if you had failover forests configured, you would still be able to access the database on the remaining two hosts.

However, if you then shut down another host, you would lose quorum (which is a requirement for failover).

Forest Availability: Local Disk Failover

For forests configured for local disk failover, the sequence of events is important:

In response to a host failure that makes an "open" forest inaccessible, the forest will failover to the configured forest replica as long as a quorum exists and the configured replica forest was in the "sync replicating" state. In this case, the configured replica forest will transition to the "open" state; the configured replica forest becomes the acting master forest and is available to the database for both reads and writes.

Additionally, an "open" forest will not go offline in response to another host being evicted from the cluster.

However, once cluster quorum is lost, forest failovers will no longer occur.

Conclusion

Depending on how your forests are distributed in the cluster and depending of the order of host failures, it is possible that a database can remain online even when there is no longer a quorum of hosts in the cluster.

Of course, databases with many forests spread across many hosts typically can't stay online if you lose quorum because some forest(s) will become unavailable.

Recommendation

Even though it is possible to have a functioning cluster with less than a quorum of hosts online, you should not architect your high availability solution to depend on it.

Summary

This article discusses what happens when you backup or restore your database after a local disk failover event on one of the database forests.

Introduction

MarkLogic Server provides high availability in the event of a data node failure. Data node failures can include operating system crashes, MarkLogic Server restarts, power failures, or persistent system failures; for example hardware failures. With Forest level failover enabled and configured, a machine that hosts a forest can go down and the MarkLogic Server cluster automatically recovers from the outage and keep continuing to process queries without any immediate action needed by an administrator. In MarkLogic Server, if a forest becomes unavailable then the entire database to which this forest is attached becomes unavailable for further query operations. Without failover, such a failure requires a manual intervention (such as administrator) to either reconfigure the forest to another host or to remove this forest from the configuration (cluster). With failover, you can configure the forest to automatically switch to a replica forest on a different host. MarkLogic Server Failover provides for high availability and maintains data and transactional integrity in the event of a data node failure.

The failover scenarios are well documented on our developer web site.

Local Disk Failover

You to configure a forest on another host to serve as a replica forest which will take over when a primary master forest's host goes offline. Local-disk failover allows you to create one or more replica forests for each primary forest. Replica forests contain the exact same data as the primary forest and are kept consistent transactionally. 

It is helpful to use the following terms to refer to the forest configurations and states:

• Configured Master is the forest which is originally configured as the primary forest.
• Configured Replica is a forest on another host that is configured as a replica forest of the primary. 
• Acting Master is the forest that is server as the master forest, regardless of the configuration.
• Acting Replica is the forest that is server as the replica forest, regardless of the configuration.

Database Backup when a forest is failed over

If you attempt to take a Database back up or perform a database restore when One of the forests of the database had failed over to the replica (i.e. Configured Replica is serving as Acting Master), it may result in XDMP-FORESTNOTOPEN or XDMP-HOSTDOWN errors.

When a database backup takes place, by default, everything associated with database gets backed up. You can also choose to backup any individual forests (only the forests selected while configuring backup are backed up). T

Replica Forest will only be backed up when the 'Include replica forests' are enabled.  If you have not configured the backup to include replica forests, then the replica forests will not be backed up even if it is the acting master. If the Configured Master is also not available, then neither forest will be backed up. In this circumstance, you may see a message in the error logs similar to "Warning: Not backing up database test because first forest master is not available, and replica backups aren't enabled."

Restore when a forest is failed over

Restore's will fail if executed when a forest is failed over (i.e. Configured Replica is serving as Acting Master). In this circumstance, you may see a message in the error logs similar to "Operation failed with error message. Check server logs." or "XDMP:HOSTDOWN".

How to detect if a forest is failed over

In the Admin UI:

1. Click the Forests icon in the left tree menu;
2. Click the Summary tab;
3. You see the configured replica in open state; (This indicates that the Configured Replica is serving as Acting Master).

At the time of the failover event, you may see messages in the Error Log similar to:
2013-10-03 12:49:53.873 Info: Disconnecting from domestic host rh6v-intel64-9.marklogic.com in cluster 16599165797432706248 because it has not responded for 30 seconds.
2013-10-03 12:49:53.873 Info: Disconnected from host rh6v-intel64-9.marklogic.com
2013-10-03 12:49:53.873 Info: Unmounted forest test_P
2013-10-03 12:49:53.875 Info: Forest test_R assuming the role of master with new precise time 13808297938747190
2013-10-03 12:49:53.875 Debug: Recovering undo on forest test_R
2013-10-03 12:49:53.875 Debug: Recovered undo at endTimestamp 13807844927734200 minQueryTimestamp 0 on forest test_R

Revert back from the failover state:

When the configured master is the acting replica, this is considered the "failover state".  In order to revert back, you must either restart the acting master forest or restart the host in which the acting master forest is locally mounted. After restarting, the forest will automatically revert to Configured Master if it's host is online. To check the status of the forests, see the Forests Summary tab in the Admin Interface. 

Conclusion 

For backup and restore to work correctly, clusters configured with local disk failover must have no forests in a failed over state. If a cluster is configured with local disk failover, and if some of its forests are failed over to their local disk replicas, the conditions causing the fail over must be resolved, and the cluster must be returned to the original forest configuration before backup and restore operations may resume.

INTRODUCTION

From the documentation:

Queries on a Replica database must run at a timestamp that lags the current cluster commit timestamp due to replication lag. Each forest in a Replica database maintains a special timestamp, called a Non-blocking Timestamp, that indicates the most current time at which it has complete state to answer a query. As the Replica forest receives journal frames from its Master, it acknowledges receipt of each frame and advances its nonblocking timestamp to ensure that queries on the local Replica run at an appropriate timestamp. Replication lag is the difference between the current time on the Master and the time at which the oldest unacknowledged journal frame was queued to be sent to the Replica.

To read more:

http://tinyurl.com/7zwq4l2

SCENARIO

Consider the following customer scenario:

• The storage the database resides on at one site fails.
• This requires the customer to run for a period of time on a single site.
• The storage / MarkLogic server are recovered at the site where the failure occurred.
• The customer needs to re-establish replication between the two sites

QUESTIONS AND ANSWERS

Q: Should we tune the lag limit to suit our application?

A: We have found in our own performance testing that increasing the lag limit beyond the default is typically not helpful.

When the master has a sustained rate of updates, a large lag limit causes it to run quickly ahead of the replica, then stall for an extended period of time until the replica catches up. This pattern repeats over and over and gives inconsistent performance on the master.

A smaller lag limit causes the master to suspend updates more frequently but for shorter periods of time, resulting in more consistent perceived performance.

Q: Is there any option to restore the replica database to a point in time from a backup of the master database & re-initiate replication from that point onwards?

A: It's fine to restore a backup to the failed system when it comes back online and before configuring replication in the reverse direction.

Q: Is there a limit to how old a backup of the replica database can be (e.g. can a replica be restored from months back in comparison to the master) and will it still sync back to the master without issue? And does this depend on what journal data is available?

A: There is no limit to how old a backup can be; the system will calculate all the deltas and apply them.

Q: Are there any documented API built-ins for any of these things?

A: Indeed; all the replication information is available through a call to xdmp:forest-status()

xdmp:forest-status( 
  xdmp:database-forests( 
    xdmp:database("MyDatabase"), 
    fn:true()))

For further information:

http://tinyurl.com/d6vbpk4

Q: Can you also advise if the replication lag limit mentioned in section 1.2.5 and the related possibility of transactions stalling on the master database applies during the bulk replication phase?

A: As long as the replica's forests are in "open replica" state, the replica will respond to queries at any commit timestamp it is able to support irrespective of whether replication is lagged.

A new feature in MarkLogic 5 is an application server setting for multi-version concurrency control (by default this is set to contemporaneous - meaning it will run from the latest timestamp that any query has committed - irrespective of whether there are still transactions in-flight).

Conversely, if nonblocking is chosen (i.e. if you create an application server to query a replica database and you set multi-version concurrency control to nonblocking), the server will choose the last timestamp where all pending transactions are known to have successfully committed.

If you wish to evaluate a query against a replica database you can use xdmp:database-nonblocking-timestamp() to determine the most current query timestamp that will not block.

Introduction

Database Replication replicates fragments/documents from a source database to a target database. You may see different database sizes (even when active fragment counts are then same) between Master and Replica Databases. This article provides overview of variables and reasons behind such observation.

Database Replication:

Database Replication operates at the forest level by copying journal frames from a forest in the Master database and replaying them on a corresponding forest in the foreign replica database. In other words, this means that when Journal frames are replayed in the replica database, the same group of documents in a single stand of the master database, does not necessarily reside in the same stand on the replica database - i.e. the distribution of fragments within stands are different between the master and replicas. 

Also, Note that Master and Replica forests can be distributed differently across hosts in each cluster. Even when they are distributed identically (Master DB forest name to Replica DB forest name) you could still see a different number stand between them.

Database Size, Deleted Fragment and Merge:

Current Database Size depends on number of factors like number of documents, index, deleted fragments in Stand etc. Deleted Fragments in any stand itself depends on Merge Policy, Background Merge process, Processing Cycle available, Linux Memory Config, Memory Usage at any given time, and application usage pattern.

Conclusion:

Master Cluster and Replica Cluster are separate entities. Although connected, they operate independently. Replica Database on target cluster provides data consistency. However how data can be spread across different stands than the corresponding master, including the retention of deleted fragments, will differ between Master and Replica Cluster. Hence you may see different sizes between Master and Replica Databases, even where the active fragments are the same.

Further Reading

• Database Replication in MarkLogic Server
• KB - Disaster Recovery
• KB - Unable to Merge all Deleted Fragments

Introduction

If your MarkLogic Server has it's logging level set to "Debug", it's common to see a chain of 'Detecting' and 'Detected' messages that look like this in your ErrorLogs:

2015-01-27 11:11:04.407 Debug: Detected indexes for database Documents: ss, fp, fcs, fds, few, fep, sln
2015-01-27 11:11:04.407 Debug: Detecting compatibility for database Documents
2015-01-27 11:11:04.407 Debug: Detected compatibility for database Documents

This message will appear immediately after forests are unmounted and subsequently remounted by MarkLogic Server. Detecting indexes is a relatively lightweight operation and usually has minimal impact on performance.

What would cause the forests to be unmounted and remounted

• Forest failovers
• Heavy network activity leading to a cluster (XDQP) "Heartbeat" timeout
• Changes made to forest configuration or indexes
• Any incident that may cause a "Hung" message

Apart from the forest state changes (unmount/mount), this message can also appear due to other events requiring index detection.

What are "Hung" messages?

Whenever you see a "Hung" message it's very often indicative of a loss of connection to the IO subsystem (especially the case when forests are mounted on network attached storage rather than local disk). Hung messages are explained in a little more detail in this Knowledgebase article:
https://help.marklogic.com/Knowledgebase/Article/View/35/0/hung-messages-in-the-errorlog

What do the "Detected" messages mean and what can I do about them?

Whenever you see a group of "Detecting" messages:

2015-01-14 13:06:26.016 Debug: Detecting indexes for database XYZ

There was an event where MarkLogic chose to (or was required to) attempt to unmount and remount forests (and the event may also be evident in your ErrorLogs).

The detecting index message will occur soon after a remount, indicating that MarkLogic Server is examining forest data to check whether any reindexing work is required for all databases available to the node which have Forests attached:

2015-01-14 13:06:26.687 Debug: Detected indexes for database XYZ: ss, wp, fp, fcs, fds, ewp, evp, few, fep

The line immediately below indicates that the scan has been completed and the database has been identified as having been configured with a number of indexes. For the line above, these are:

ss

stemmed searches

wp

word positions

fp

fast phrase searches

fcs

fast case sensitive searches

fds

fast diacritic sensitive searches

ewp

element word positions

evp

element value positions

few

fast element word searches

fep

fast element phrase searches

From this list, we are able to determine which indexes were detected.  These messages will occur after every remount if you have index detection set to automatic in the database configuration.

Every time the forest is remounted, in addition to a recovery process (where the Journals are scanned to ensure that all transactions logged were safely committed to on-disk stands), there are a number of other tests the server will do. These are configured with three options at database level:

• format compatibility
• index detection
• expunge locks

By default, these three settings are configured with the "automatic" setting (in MarkLogic 7), so if you have logging set to "Debug" level, you'll know that these options are being worked through on remount:

2015-01-14 13:06:26.016 Debug: Detecting indexes for database XYZ (represents the task for "automatic" index detection where the reindexer checks for configuration changes)
2015-01-14 13:06:26.687 Debug: Detecting compatibility for database XYZ (represents the task for "automatic" format compatibility where the on-disk stand format is detected)

These default values may change in accross releases of MarkLogic Server. In MarkLogic 8, expunge locks is set to none but the other two are still set to automatic.

Can these values be changed safely and what happens if I change these?

Unmounting / remounting times can be made much shorter by configuring these settings away from automatic but there are some caveats involved; if you need to upgrade to a future release of the product, it's likely that the on-disk stand format may change (it's still 5.0 even when MarkLogic 8 is released) and so setting format compatibility to 5.0 should cause the "Detecting compatibility" messages to disappear and speed up remount times.

The same is true for disabling index detection but it's important to note that changing index settings on the database will no longer cause the reindexer to perform any checks on remount; in this case you would need to enable this for changes to database index settings to be reindexed.

Related Reading

How to handle XDQP-TIMEOUT on a busy cluster

Summary

This article will provide steps to debug applications using the Alerting API that are not triggering an alert.

Details

1) Check that all required components are present in the database where alerting is setup: config, actions, rules.   Run the attached script 'getalertconfigs.xqy' through the Query Console and review the output.  

2) As documented in our Search Developer's Guide, Test the alert manually with alert:invoke-matching-actions(). 

Example:

alert:invoke-matching-actions("my-alert-config-uri", 
      <doc>hello world</doc>, <options/>)

3) Use the rule's query to test against the database to check that the expected documents are returned by the query.

Take the query text from the rule and run it through Query Console using a cts:search() on the database.  This will confirm whether the expected documents are a positive match.  If the documents are returned and no alert is triggered, then further debugging will be needed on the configuration or the query may need to be modified.

Introduction 

Division operations involving integer or long datatypes may generate XDMP-DECOVRFLW in MarkLogic 7. This is the expected behavior but it may not be obvious upon initial inspection.  

For example, similar queries with similar but different input values executed in Query Console on Linux/Mac machine running MarkLogic 7 gives the following results

1. This query returns correct results

let $estimate := xs:unsignedLong("220")

let $total := xs:unsignedLong("1600")

return $estimate div $total * 100

==> 13.75

2. This query returns the XDMP-DECOVRFLOW Error

let $estimate := xs:unsignedLong("227")

let $total := xs:unsignedLong("1661")

return $estimate div $total * 100

==> ERROR : XDMP-DECOVRFLW: (err:FOAR0002)

Details

The following defines relevant behaviors in MarkLogic 7 and previous releases.

• In MarkLogic 7, if all the operands involved in div operations are integer, long or integer sub-types in XML, then the resulting value of the div operation are stored as xs:decimal.
• In versions previous to MarkLogic 7, if an xs:decimal value is large and occupies all digits then it was implicitly cast into an xs:double for further operations - i.e. beginning with MarkLogic, implict casting no longer occurs in this situation .
• xs:decimal can accomodate 18 digits as a datatype.
• In MarkLogic 7 on Linux & Mac, xs:decimal can occupy all digits depending upon actual value ( 227 div 1661 = 0.1366646598434677905 ), all 18 digits occupied in xs:decimal
• MarkLogic 7 on Windows does not perform division with full decimal precision ( 227 div 1661 produces 0.136664659843468 ); as a result, not all 18 digits occupied in xs:decimal
• MarkLogic 7 will generates Overflow Exception : FOAR0002, when an operation is performed on an xs:decimal that is already at full decimal precision

In the example above, multiplying the result with 100 gives an error in Linux/Mac, while its OK on Windows.

Recommendations:

We recommend xs:double be used for all division related operations in order to explicitly cast resulting value to larger data-type.

For example: These will return results

xs:double($estimate) div $total * 100


$estimate div $total * xs:double(100)

.

Context:

There are options 'maintain last modified' and 'maintain directory last modified' on the Admin UI for a database, which when turned on add properties to every document inserted in the database.  There may be a need to remove all the property fragments of all the documents in the database when the properties no longer need to be retained.

Problem:

Turning these options off for a database ensure that properties will not be created for new documents. However, existing document properties will not be removed by turning these settings off.

Solution:

To delete existing document properties, the following query can be used:

xdmp:node-delete(xdmp:document-properties(“your-document-uri”))

Please make sure that 'maintain last modified' and 'maintain directory last modified' options are turned off for the database, so that the property fragment does not get recreated for the document.

Summary

Terraform from HashiCorp is a deployment tool that many organizations use to manage their infrastructure as code. It is platform agnostic, allowing for the deployment and configuration of on-site physical infrastructure, as well as cloud infrastructure such as AWS, Azure, VSphere and more.

Terraform uses the Hashicorp Configuration Language (HCL) to allow for concise descriptions of infrastructure. HCL is JSON compatible language, and was designed to be both human and machine friendly.

This powerful tool can be used to deploy a MarkLogic Cluster to AWS using the MarkLogic CloudFormation Template. The MarkLogic CloudFormation Template is the preferred method recommended by MarkLogic for building out MarkLogic clusters within AWS.

Setting Up Terraform

For the purpose of this example, I will assume that you have already installed Terraform, the AWS CLI and you have configured the credentials. You will also need to have a working directory that has been initialized using terraform init.

Terraform Providers

Terraform uses Providers to provide access to different resources. The Provider is responsible for understanding API interactions and exposing resources. The AWS Provider is used to provide access to AWS resources.

Terraform Resources

Resources are the most important part of the Terraform language. Resource blocks describe one or more infrastructure objects, like compute instances and virtual networks.

The aws_cloudformation_stack resource, allows Terraform to create a stack from a CloudFormation template.

Choosing a Template

MarkLogic provides two templates for creating a managed cluster in AWS.

• MarkLogic cluster in new VPC
• MarkLogic cluster in an existing VPC

Defining Variables

The MarkLogic CF Template takes a number of input variables, including the region, availability zones, instance types, EC2 keys, encryption keys, licenses and more. We have to define our variables so they can be used as part of the resource.

Variables in HCL can be declared in a separate file, which allows for deployment flexibility. For instance, you can create a Development resource and a Production resource, but using different variable files.

Here is a snippet from our variables file:

variable "cloudform_resource_name" {
type = string
default = "Dev-Cluster-CF"
}
variable "stack_name" {
type = string
default = "Dev-Cluster"
}
variable "ml_version" {
type = string
default = "10.0-4"
}
variable "availability_zone_names" {
type = list(string)
default = ["us-east-1a","us-east-1b","us-east-1c"]
}
...

In the snippet above, you'll notice that we've defined the availability_zone_names as a list. The MarkLogic CloudFormation template won't take a list as an input, so later we will join the list items into a string for the template to use.

This also applies to any of the other lists defined in the variable files.

Using the CloudFormation Resource

So now we need to define the resource in HCL, that will allow us to deploy a CloudFormation template to create a new stack.

The first thing we need to do, is tell Terraform which provider we will be using, defining some default options:

    provider "aws" {
    profile = "default"
    #access_key = var.access_key
    secret_key = var.secret_key
    region = var.aws_region
    }

Next, we need to define the `aws_cloudformation_stack` configuration options, setting the variables that will be passed in when the stack is created:

    resource "aws_cloudformation_stack" "marklogic" {
    name = var.cloudform_resource_name
    capabilities = ["CAPABILITY_IAM"]


    parameters = {
    IAMRole = var.iam_role
    AdminUser = var.ml_admin_user
    AdminPass = var.ml_admin_password
    Licensee = "My User - Development"
    LicenseKey = "B581-REST-OF-LICENSE-KEY"
    VolumeSize = var.volume_size
    VolumeType = var.volume_type
    VolumeEncryption = var.volume_encryption
    VolumeEncryptionKey = var.volume_encryption_key
    InstanceType = var.instance_type
    SpotPrice = var.spot_price
    KeyName = var.secret_key
    AZ = join(",","${var.avail_zone}")
    LogSNS = var.log_sns
    NumberOfZones = var.number_of_zones
    NodesPerZone = var.nodes_per_zone
    VPC = var.vpc_id
    PublicSubnets = join(",","${var.public_subnets}")
    PrivateSubnets = join(",","${var.private_subnets}")
    }
    template_url = "${var.template_base_url}${var.ml_version}/${var.template_file_name}"
    }

Deploying the Cluster

Now that we have defined our variables and our resources, it's time for the actual deployment.

$> terraform apply

This will show us the work that Terraform is going to attempt to perform, along with the settings that have been defined so far.

Once we confirm that things look correct, we can go ahead and apply the resource.

Now we can check the AWS Console to see our stack

And we can also use the ELB to login to the Admin UI

Wrapping Up

We have now deployed a 3 node cluster to an existing VPC using Terraform. The cluster is now ready to have our Data Hub, or other application installed.

Deploying REST API Search/Query Options in DHS

REST API Query Options Overview

You can use persistent or dynamic query options to customize your queries. MarkLogic Server comes configured with default query options. You can extend and modify the default options using /config/query/default.

REST API Search options are defined per Group and App Server. When using ml-gradle, they are typically deployed by putting the files defining the options in the src/main/ml-modules/options directory of your gradle project. By default the options will be deployed to the Group/App Server that gradle is pointing at in the data-hub-MODULES database under /[GroupName]/[App Server]/rest-api/options/[name of file].

REST API Query Options in DHS

In DHS, query options are created under the Evaluator Group for the data-hub-FINAL app server. One side effect of the permissions for DHS, is that users will not be able to see the files after they are deployed. The default permissions for the options file are rest-reader-internal and rest-admin-internal, which is not provided to the data-hub roles.

To check that the search options have been deployed you can use the following curl command:

• curl --anyauth --user username:password -k -X GET -H "Content-type: application/xml" https://myService.a.marklogicsvc.com:8011/v1/search?options=[myOptions]

If the options exist, you will get results. If the options do not exist, then you will get a 400 return, with a REST-INVALIDPARAM error.

Deploying Options to Other App Servers and Groups

Deploying Options to the Staging App Server

Using src/main/ml-modules/options will only deploy the options to the Final app server. If you want to deploy the options to the Staging app server, then you will need to define the options under src/main/ml-modules/root/Evaluator/data-hub-STAGING/rest-api/options

Deploying Options to Other Groups

If the cluster is configured for auto-scaling, the dynamic e-nodes will belong to either the Analyzer, Curator or Operator group, so the search options will not be available for the dynamic e-nodes.

To set the options for the app servers in other groups, you will also use src/main/ml-modules/root/[Group Name]/[App Server Name]/rest-api/options

• src/main/ml-modules/root/Analyzer/data-hub-FINAL/rest-api/options
• src/main/ml-modules/root/Operator/data-hub-FINAL/rest-api/options
• ...etc

When deploying the options files in this way, they get different permissions than when they are deployed vi ml-modules/options. The permissions are rest-extension-user, data-hub-module-reader, data-hub-module-writer, tde-admin, and tde-view, but the permission differences do not appear to make a difference in functionality.

Deployment Failures

When options are deployed with the rest of the non-REST modules in ml-modules/root/..., it uses the /v1/documents endpoint, which allows you to set the file permissions.

When options are deployed from ml-modules/options, it uses the /v1/config/query endpoint, which does not allow you to set the file permissions.

One effect of this difference is if you attempt to deploy the search options using both ml-modules/options and src/main/ml-modules/root/Evaluator/data-hub-FINAL/rest-api/options you will encounter a SEC-PERMDENIED error and the deployment will fail. If you encounter this error, ensure you aren't attempting to deploy the options in both locations.

Introduction

This KB article lists some available tools for continuous integration and automatically deploying the MarkLogic Server

Deployment Options

ml-gradle is a gradle plugin that can be used for configuration and application deployments. Application deployments are maintained as projects, which can deployed to any environment - Development, QA, Production, etc.

The MarkLogic Configuration Management API is a RESTful API that allows retrieving, generating, and applying configurations for MarkLogic clusters, databases, and application servers.

The MarkLogic The Management API is a REST-based API that allows you to administer MarkLogic Server and access MarkLogic Server instrumentation with no provisioning or set-up. You can use the API to perform administrative tasks such as initializing or extending a cluster; creating databases, forests, and App Servers; and managing tiered storage partitions. The API also provides the ability to easily capture detailed information on MarkLogic Server objects and processes, such as hosts, databases, forests, App Servers, groups, transactions, and requests from a wide variety of tools.

The MarkLogic Admin APIs provide a flexible toolkit for creating new and managing existing configurations of MarkLogic Server.

Integration Testing

MarkLogic Unit Test is a testing component that was originally part of the Roxy project. This component enables you to build unit tests that are written in and can test against both XQuery and Server-side JavaScript.

Implementation Specific Tools

CloudFormation Templates

MarkLogic CloudFormation templates enable you to launch clusters with an Elastic Load Balancer, Elastic Block Storage, Auto Scaling Group, and so on. Your cluster can be in either one Availability Zone or three Availability Zones. Multiple nodes can be placed within each Availability Zone. You can choose whether to deploy to an existing VPC, or a new VPC. The templates can also be used with tools like Terraform and Ansible

• Deploying MarkLogic in AWS with Terraform
• Deploying MarkLogic to AWS with Ansible

Python

The MarkLogic Python API aims to provide complete coverage of the capabilities in the MarkLogic REST API in idiomatic Python.

Jenkins

Jenkins is often used with MarkLogic Server for building deployable artifacts, staging build artifacts, running automated tests, and deploying said artifacts. Jenkins has great REST endpoints that make it easy to get / put job configurations, and enable / disable jobs from scripts.

Jenkins provides a driver to the continuous integration / continuous delivery process that can integrate with other tools. In combination with ml-gradle, it can be used to run deploy module/unit test on code check-in.

One pipeline example used with Jenkins is to:

1. Pull the code from Git
2. Deploy to DEV with ml-gradle
3. Run MarkLogic Unit Test
4. Email a report of the success/failure
5. Kick off job to deploy to another environment

Also noted that the most important best practice here would be to make sure Jenkins runs primarily off of a host other than a MarkLogic host.

SUMMARY

This article will help MarkLogic Administrators to monitor the health of their MarkLogic cluster. By studying the attached scripts, you will learn how to find out which hosts are down and which forests have failed over, enabling you to take the necessary recovery actions.

Initial Setup

On a separate Linux host (not a member of the cluster), download the file attachments from this article, making sure that they all reside within the same directory.

Here is a general description of each file:

cluster-name.conf - Example configuration file used by script. Configures information for monitoring one ML cluster. 

ml-ck-for-life.sh - A very simple, low-load check that all the nodes of a cluster are up and running.

ml-ck-for-health.sh - A more detailed check for essential cluster functionality with alerting (paging and/or emails to DBAs) if warranted. This script relies on at least one external XQuery file (mon-report-failed-over-forests.xqy) and makes use of the REST MGMT API as well as REST XQuery requests.

mon-report-failed-over-forests.xqy - External XQuery file used by ml-ck-for-health.sh

Preparing the CONF File for Use on Your Cluster

Before running the scripts, the cluster-name.conf needs to be customized for your specific cluster. Start by changing the file name to match the name of your cluster, e.g.,

$ mv cluster-name.conf some-other-name.conf

Where "some-other-name" is the actual name of the cluster, or of the application that is hosted on that cluster.

Next, you will need to customize some of the internal variables inside the CONF file itself. Here is the contents of the cluster-name.conf file, as downloaded:

CLUSTER_NAME="CLUSTER-NAME"
CLUSTER_NODES=( node1.my-company.com node2.my-company.com node3.my-company.com )
# MarkLogic Credentials for the REST Management port - 8002
USER_PW_MGMT=rest-manager-user:re-manager-password
# MarkLogic Credentials for the XQuery eval port - 8000
USER_PW_XQ=user-name:user-password
UNIX_USER=unix-user-name
PAGE_ADDRESSES=ml.alert.page@my-company.com
MAIL_ADDRESSES=ml.alert.mail@my-company.com

---------  end of listing ---------

For CLUSTER_NAME, provide the cluster-name listed in the cluster's /var/log/MarkLogic/clusters.xml file.

For CLUSTER_NODES, write in the host-names for each node in your cluster.

For USER_PW_MGMT, provide the user-name and password for the REST MANAGEMENT user, the format is name:password.

For USER_PW_XQ, provide the user-name and password for the user who will execute the XQuery scripts, the format is name:password.

The UNIX_USER is a local Unix username with the correct rwx access rights for this directory.

The PAGE_ADDRESSES & MAIL_ADDRESSES are alert email addresses who will be notified whenever there is a failover event.

Periodicity

The script ml-ck-for-health.sh was created with the idea it would be run repeatedly at a certain interval to keep tabs on system health. For example, it can be configured to be invoked with a cron job. A frequency of 5 to 120 minutes is a good candidate range. Ten minutes is a good time if you would like to be woken up (on average) within 5 minutes of a failover event.

Setting up SSH Passwordless Login

In monitoring script ml-ck-for-health.sh, section (6) FOREST STATUS CHANGE, requires ssh access to the cluster hosts. That is because this section greps through MarkLogic server ErrorLogs. To enable this part of the script to run without prompting the user, "ssh passwordless login" should be setup between the monitoring host and all the cluster hosts.There are many examples of how to do this on the internet, for example: http://www.tecmint.com/ssh-passwordless-login-using-ssh-keygen-in-5-easy-steps/ Alternatively, this monitoring section can be commented out.

Also regarding section (6), the “grep” command is setup up to grep the latest 10 minutes from the ErrorLog. If this script is configured to be run less often then every 10 minutes, the “grep” command line should be adapted to cover the desired period between script runs.

Example Usage

You are now ready to execute the failover monitoring scripts! Here is how you would execute them:

$ ./ml-ck-for-health.sh some-other-name.conf MY-CLUSTER-NAME

$ ./ml-ck-for-life.sh some-other-name.conf

[where "some-other-name" and MY-CLUSTER-NAME are your actual CONF and cluster-name, as described above]

Monitoring Multiple Clusters

So, given a monitoring machine with a directory of cluster configuration files in the style of cluster-name.conf, those configuration files could be iterated through to monitor a suite of clusters from a single monitoring machine. It should be fairly easy to build a custom shell script to iterate through various cluster CONF files.

Final thought and Limitations

Please be aware that the ml-ck-for-health.sh script is only partially implemented. In particular, the Replication Lag and Replication Failure sections are left as exercises for the user.

This script is being presented as a backup, lowest common denominator monitoring solution. For a more complete solution, you should explore other options, such as Splunk or Nagios.

Introduction

According to Wikipedia, DevOps is a set of practices that combines software development (Dev) and IT operations (Ops) with the goal of shortening the Systems Development Lifecycle, and providing continuous delivery with high software quality. This KB will provide some guidance for system deployment and configuration, which can be integrated into an organizations DevOps processes.

For more information on using MarkLogic as part of a Continuous Integration/Continuous Delivery process, see the KB  Deployment and Continuous Integration Tools.

Deploying a Cluster

Deploying a MarkLogic cluster that will act as the target environment for the application code being developed is one piece of the DevOps puzzle. The approach that is chosen will depend on many things, including the tooling already in use by an organization, as well as the infrastructure that will be used for the deployment.  We will cover two of the most common environments, On-Premise and Cloud.

On-Premise Deployments

On-Premise deployments, which can include using bare metal servers, or Virtual Machine infrastructure (such as VMware), are one common environment. You can deploy a cluster to an on-premise environment using tools such as shell scripts, or Ansible. In the Scripting Administrative Tasks Guide, there is a section on Scripting Cluster Management, which provides some examples of how a cluster build can be automated.

Once the cluster is deployed, some of the specific configuration tasks that may need to be performed on the cluster can be done using the Management API.

Cloud Deployments

Cloud deployments utilize flexible compute resources provided by vendors such as Amazon Web Services (AWS), or Microsoft Azure.

For AWS, MarkLogic provides an example CloudFormation template, that can be used to deploy a cluster to Amazon's AWS EC2 Environment. Tools like the AWS Command Line Interface (CLI), Terraform or Ansible can be used to extend the MarkLogic CloudFormation template, and automate the process of creating a cluster in the AWS EC2 environment.  MarkLogic has provided an example , which can be utilized to . The template can be used to deploy a cluster using the AWS CLI. The template can also be used to Deploy a Cluster Using Terraform, or it can be used to Deploy a Cluster Using Ansible.

For Azure, MarkLogic has provided Solution Templates for Azure which can be extended for automated deployments using the Azure CLI, Terraform or Ansible.

As with the on-premise deployments, configuration tasks can be performed on the cluster using the Management API

Summary

This is just a brief introduction into some aspects of DevOps processes for deploying and configuring a MarkLogic Cluster.

Summary:

After adding or removing a forest and correspond replica forest in a database, we have seen instances where the Rebalancer does not properly distribute the documents amongst existing and newly added forests.

For this particular instance, XDMP-HASHLOCKINGRETRY debug level error message reported repeatedly in the error logs.  The messages would look something like: 

2016-02-11 18:22:54.044 Debug: Retrying HTTPRequestTask::handleXDBCRequest 83 because XDMP-HASHLOCKINGRETRY: Retry hash locking. Forests config hash does not match.

2016-02-11 18:22:54.198 Debug: Retrying ForestRebalancerTask::run P_initial_p2_01 50 because XDMP-HASHLOCKINGRETRY: Retry hash locking. Forests config hash does not match.

Diagnosing

Gather statistics about the rebalancer to see the number of documents being scheduled. If you run attached script “rebalancer-preview.xqy” in the query console of your MarkLogic Server cluster, it will produce rebalancer statistics in tabular format.

• Note that you must first change the database name (YourDatabaseNameOnWhichNewForestsHaveBeenAdded) on the 3^rd line of the XQuery script “rebalancer-preview.xqy”:

declare variable $DATABASE as xs:string := xdmp:get-request-field("db", "YourDatabaseNameOnWhichNewForestsHaveBeenAdded");

If experiencing this issue, the newly added forests will show zero in the “Total to be moved” column in the generated html page.

Resolving

Perform a cluster wide restart in order to get past this issue.  The restart is required to reload all of the configuration files across the cluster.  The rebalancer will also check to see if additional rebalancing work needs to occur. The rebalancer should work as expected now and the  XDMP-HASHLOCKINGRETRY messages should no longer appear in the logs. If you run the rebalancer-preview.xqy script again, the statistics should now show the the number of documents being scheduled to be moved.

You can also validate the rebalancer status from the Database Status page in the Admin UI.

The XDMP-HASHLOCKINGRETRY rebalancer issue has fixed in the latest MarkLogic Server releases.  However, the rebalancer-preview.xqy script can be used to help diagnose other perceived issues with the Rebalancer.

Search fundamentals

Difference between cts:contains and fn:contains

 1) fn:contains is a substring match, where as cts:contains performs query matching

 2) cts:contains therefore can utilize general queries and stemming, where fn:contains does not

For example:-

Example.xml

<test>daily running makes you fit</test>

•         fn:contains(fn:doc(“Example.xml”),”ning”)

          True

•          cts:contains(fn:doc(“Example.xml”),”ning”)

         False

•         fn:contains(fn:doc(“Example.xml”),”ran”)

           False

•         cts:contains(fn:doc(“Example.xml”),”ran”)

            True

Note:-

The cts:contains examples are checking the document against cts:word-querys.  Stemming reduces words down to their root, allowing for smaller term lists.

1) Words from different languages are treated differently, and will not stem to the same root word entry from another language.

2) Note: Nouns will not stem to verbs and vice versa. For example, the word “runner” will not stem to “run”.

References

• fn:contains
• cts:contains

Introduction

MarkLogic Server provides a variety of  disaster recovery (DR) facilities including full backup, incremental backup, and journal archiving that when combined with other ML features can create a complete disaster recovery strategy. This paper shows some examples of how these features can be combined. It is not comprehensive nor does it reflect features offered only in the latest releases.

Details

This article will cover three perspectives. First, a quick overview of the metrics used by businesses to measure the quality of their Disaster Recovery strategies will be covered. Next, an overview of how to combine the features that MarkLogic offers in various categories will be given.

More?: High Availability and Disaster Recovery features ,  High Availability & Disaster Recovery datasheet, Scalability, Availability, and Failover Guide 

Disaster Recovery Criteria

In order to configure MarkLogic Server to perform well in Disaster Recovery situations, we should first define what parameters we will use to measure each possible approach. For most situations, these four measures are used: 

Long Term Retention Policy (LTR): Long Term Retention Policy can be driven by any number of business, regulatory and other criteria. It is included here because MarkLogic's backup files are often a key part of an LTR strategy. 

Recovery Point Objective (RPO): The requirement for how up-to-date the database has to be post-recovery with respect to its state immediately before the incident that required recover.

Recovery Time Objective (RTO): The requirement for the time elapsed between the incident and the recovery to the RPO.

Cost: The storage cost, the computational resource cost and  the operations cost of the overall deployment strategy.

Flexible Replication Features

Flexible replication can be used to support LTR objectives but is generally not useful for Disaster Recovery

More? Flexible Replication Guide

Platform Support Features

Flash backup provides a way to leverage backup features of your deployment platform while maintaining transaction integrity. Platform specific solutions can often achieve RPO and RTO targets that would be impossible through other means.

More? Flash Backup

High Availability Features

Forest replication provides recovery from host failures.

More? Scalability, Availability, and Failover Guide

Disaster Recovery Features

Database Replication

Database Replication is the process of maintaining copies of forests on databases in multiple MarkLogic Server clusters.

More? Understanding Database Replication

Backups

Of all your backup options, full backups restore the quickest, but take the most time to backup and possibly the most storage space. Each full backup is a backup set in that it contains everything you need to restore to the point of the backup.

Full backups with journal archiving allow restores to a point after the backup, but the journal archive grows in an unbounded way with the number of transactions, and replaying the journals to get to your recovery point takes time proportional to the number of transactions in the journal archive, so over time, this becomes less efficient.

With full + incremental backups, a backup set is a full backup, plus the incremental backups taken after that full backup. Incremental backups are quick to backup, but take longer to restore, and over time the backup set gets larger and larger, so it may end up consuming more backup space than a full backup alone (depending on your backup retention policy).

Full + incremental backups with journal archiving have the same characteristics as incremental backups, except that you can roll forward from the most recent incremental. With this strategy, the journal archive doesn't grow in an unbounded way because the archive is purged when you take the next incremental backup. Note that if your RPO is between incremental backups, you must also enable a merge timestamp by setting the merge timestamp to a negative value (see below).

More?: Administrator’s Guide to Backing Up and Restoring a Database  How does "point-in-time" recovery work with Journal Archiving? 

Forest Merge Configurations

Forest merges recover the disk space occupied by deleted documents. A negative merge timestamp delays that permanent deletion. If we want incremental backups to contain all the fragments that were deleted since the last incremental backup then we want to set the delay to a period greater than the incremental backup period. This requires more disk space for the incremental backups and also requires additional space in the live database, but provides the most flexibility.

Setting retain-until-backup on a given database (thru the Admin UI or thru an API call) has a similar effect by telling the server to keep the deleted fragments until a full backup or an incremental backup completes. Many clients choose to use both the negative merge timestamp and retain until backup options together.

More?: admin:database-set-merge-timestamp  admin:database-set-retain-until-backup

Other Features

The need for a negative merge timestamp can be understood by remembering that forest merges recover the disk space occupied by deleted documents. A negative merge timestamp delays that permanent deletion. If we want incremental backups to contain all the fragments that were deleted since the last incremental backup then we want to set the delay to a period greater than the incremental backup period. This requires more disk space for the incremental backups and also requires additional space in the live database, but provides the most flexibility.

Setting retain-until-backup on a given database (thru the Admin UI or thru an API call) has a similar effect by telling the server to  keep the deleted fragments until a full backup or an incremental backup. Many clients choose to use both the negative merge timestamp and retain until backup options together.

More?: admin:database-set-merge-timestamp,  admin:database-set-retain-until-backup 

Conclusion

Planning to meet a Long Term Retention (LTR) policy, a Recovery Point Objective (RPO) and a Recovery Time Objective (RTO) and a Cost goal is a key part of developing an overall MarkLogic deployment plan. MarkLogic offers a wealth of tools that can complement each other when they are properly coordinated. As is clear from this article, the choices are many, broad, and interrelated.

Regardless of the server version, MLCP does not support concurrent jobs if they are importing from/exporting to the same file.

In general, MLCP jobs will perform best by maximizing the number of threads in a single MLCP job. Before 10.0-4.2, each MLCP job used 4 threads by default. Starting in 10.0-4.2, each MLCP job now uses the maximum number of threads available on the server as the default thread count (you can read more about this change in the 10.0-4.2 release notes).

Introduction

In the more recent versions of MarkLogic Server, there are checks in place to prevent the loading of invalid documents (such as documents with multiple root nodes).  However, documents loaded in earlier versions of MarkLogic Server can now result in duplicate URI or duplicate document errors being reported.

Additionally, under normal operating conditions, a document/URI is saved in a single forest. If somehow the load process gets compromised, then user may see issues like duplicate URI (i.e. same URI in different forests) and duplicate documents (i.e. same document/URI in same forest).

Resolution

If the XDMP-DBDUPURI (duplicate URI) error is encountered, refer to our KB article "Handling XDMP-DBDUPURI errors" for procedures to resolve.

If one doesn't see XDMP-DBDUPURI errors but running fn:doc() on a document returns multiple nodes then it could be a case of duplicate document in same forest.

To check that the problem is actually duplicate documents, one can either do an xdmp:describe(fn:doc(...)) or fn:count(fn:doc((...)). If these commands return more than 1 e.g. xdmp:describe(fn:doc("/testdoc.xml")) returns (fn:doc("/testdoc.xml"), fn:doc("/testdoc.xml")) or fn:count(fn:doc("/testdoc.xml")) returns 2 then the problem is of duplicate documents in the same forest (and not duplicate URIs).

To fix duplicate documents, the document will need to be reloaded.

Before reloading, you can take a look at the two version to see if there is a difference.  Check fn:doc("/testdoc.xml")[1] versus fn:doc("/testdoc.xml")[2] to see if there is a difference, and which one you want to reload.

If there is a difference, that may also that may point the operation that created the situation.

Introduction

This article talks about effects of case sensitivity of search term on search score and thus on final order of search results for a secondary query which is using cts:boost-query and weight. The case-insensitive word term is treated as the lower case word term, so there can be no difference in the frequencies and scores of results for any-case/case-insensitive search term and lowercase search term with “case-sensitive” option or when neither "case-sensitive" nor "case-insensitive" is present. If neither "case-sensitive" nor "case-insensitive" is present, text of search term is used to determine case sensitivity.

Understanding relevance score

In MarkLogic Search results are returned in a relevance order. The most relevant results are first in result sequence and least relevant are last.
More details on relevance score and its calculation are available at, https://docs.marklogic.com/guide/search-dev/relevance

Of many ways to control this relevance score one way is to use a secondary query to boost relevance score, https://docs.marklogic.com/guide/search-dev/relevance#id_30927 . This article takes advantage of examples using secondary query to boost relevance scores and impact of text case (upper, lower or unspecifed) of search terms on relevance score on order of results returned.

A few examples to understand this scenario

Consider a few scenarios where below mentioned queries are trying to boost certain search results up using cts:boost-query and weight for word "washington" in returned results.

Example 1: Search with lowercase search term and option for case not specified

Query1:
xquery version "1.0-ml";
declare namespace html = "http://www.w3.org/1999/xhtml";

for $hit in
( cts:search(
fn:doc()/test,

cts:boost-query(cts:element-word-query(xs:QName("test"),"George" ),
cts:element-word-query(xs:QName("test"),"washington",(), 10.0) )
)
)

return element hit {
attribute score { cts:score($hit) },
attribute fit { cts:fitness($hit) },
attribute conf { cts:confidence($hit) },
$hit
}

Results for Query1:
<hit score="28276" fit="0.9393904" conf="0.2769644">
<test>Washington, George... </test>
</hit>
...
...
<hit score="16268" fit="0.7125317" conf="0.2100787">
<test>George washington was the first President of the United States of America...</test>
</hit>
...

Example 2: Search with lowercase search term and case-sensitive option

Query2:
xquery version "1.0-ml";
declare namespace html = "http://www.w3.org/1999/xhtml";

for $hit in
( cts:search(
fn:doc()/test,

cts:boost-query(cts:element-word-query(xs:QName("test"),"George" ),
cts:element-word-query(xs:QName("test"),"washington",("case-sensitive"), 10.0) )
)
)

return element hit {
attribute score { cts:score($hit) },
attribute fit { cts:fitness($hit) },
attribute conf { cts:confidence($hit) },
$hit
}

Results for Query2:
<hit score="28276" fit="0.9393904" conf="0.2769644">
<test>Washington, George... </test>
</hit>
...
...
<hit score="16268" fit="0.7125317" conf="0.2100787">
<test>George washington was the first President of the United States of America...</test>
</hit>
...

Example 3: Search with uppercase search term and option case-insensitive, in cts:boost-query like below with rest of query similar to above queries

Query3:

cts:boost-query(cts:element-word-query(xs:QName("test"),"George" ),
cts:element-word-query(xs:QName("test"),"Washington",("case-insensitive"), 10.0) )

Results for Query3:
<hit score="28276" fit="0.9393904" conf="0.2769644">
<test>Washington, George... </test>
</hit>
...
...
<hit score="16268" fit="0.7125317" conf="0.2100787">
<test>George washington was the first President of the United States of America...</test>
</hit>
...

Clearly above queries are producing the same scores with same fitness and confidence scores. This is because the case-insensitive word term is treated as the lower case word term, so there can therefore be no difference in the frequencies of those two terms (any-case/case-insensitive and lowercase/case-sensitive), and therefore no difference in scoring. Thus no difference in scores of results for Query3 and Query2.
And for cases where case sensitivity is not specified, text of search term is used to determine case sensitivity. For Query3 text of search term contains no uppercase hence it treated as "case-insensitive".

Now let us now take look at a query with a word with uppercase and case-sensitive option in query.

Example 4: Search with uppercase search term and option case-sensitive, in cts:boost-query like below with rest of query similar to above queries

Query4:

cts:boost-query(cts:element-word-query(xs:QName("test"),"George" ),
cts:element-word-query(xs:QName("test"),"Washington",("case-sensitive"), 10.0) )

Results for Query4:
<hit score="44893" fit="0.9172696" conf="0.3489831">
<test>Washington, George was the first... </test>
</hit>
...
...
<hit score="256" fit="0.0692672" conf="0.0263533">
<test>George washington was the first President of the United States of America...</test>
</hit>
...

As we can clearly see the scores are changed for results for Query4 and thus final order of results is also updated.

Conclusion:

While using a secondary query having cts:boost-query and weight, to boost certain search results up, it is important to understand the impact of case of search text on result sequence. A case-insensitive word term is treated as the lower case word term, so there can therefore be no difference in the frequencies of any-case/case-insensitive and lowercase/case-sensitive search terms, and therefore no difference in scoring. For search term with upper case alphabets in text and with “case-sensitive” option scores are boosted up as expected in comparison with a “case-insensitive search”. If neither "case-sensitive" nor "case-insensitive" is present, text of search term is used to determine case sensitivity. If text of search term contains no uppercase, it specifies "case-insensitive". If text of search term contains uppercase, it specifies "case-sensitive".

Background

MarkLogic Server includes element level security (ELS), an addition to the security model that allows you to specify security rules on specific elements within documents. Using ELS, parts of a document may be concealed from users who do not have the appropriate roles to view them. ELS can conceal the XML element (along with properties and attributes) or JSON property so that it does not appear in any searches, query plans, or indexes - unless accessed by a user with appropriate permissions.

ELS protects XML elements or JSON properties in a document using a protected path, where the path to an element or property within the document is protected so that only roles belonging to a specific query roleset can view the contents of that element or property. You specify that an element is part of a protected path by adding the path to the Security database. You also then add the appropriate role to a query roleset, which is also added to the Security database.

ELS uses query rolesets to determine which elements will appear in query results. If a query roleset does not exist with the associated role that has permissions on the path, the role cannot view the contents of that path.

Notes:

1. A user with admin privileges can access documents with protected elements by using fn:doc to retrieve documents (instead of using a query). However, to see protected elements as part of query results, even a user with admin privileges will need to have the appropriate role(s).
2. ELS applies to both XML elements and JSON properties; so unless spelled out explicitly, 'element' refers to both XML elements and JSON properties throughout this article.

You can read more about how to configure Element Level Security here, and can see how this all works at this Element Level Security Example.

Node-update

One of the commonly used document level capabilities is 'update'. Be aware, however, that document level update is too powerful to be used with ELS permissions as someone with document level update privileges could update not only a node, but also delete the whole document. Consequently, a new document-level capability - 'node-update' - has been introduced. 'node-update' offers finer control when combined with ELS through xdmp:node-replace and xdmp:node-delete functions as they can be used to update/delete only the specified nodes of a document (and not the document itself in its entirety).

Document-level vs Element-level security

Unlike at the document-level:

• 'update' and 'node-update' capabilities are equivalent at the element-level. However, at the document-level, if a user only has a 'node-update' privilege to a document, you cannot delete the document. In contrast, 'update' privileges allows that user to delete the document
• 'Read', 'insert' and 'update' are checked separately at the element level i.e.:
  • read operations - only permissions with 'read' capability are checked
  • node update operations - only permissions with 'node-update' (update) capability are checked
  • node insert operations - only permissions with  'insert' capability are checked

Note: read, insert, update and node-update can all be used at the element-level i.e., they can be part of the protected path definition.

Permissions:

Document-level:

1. update: A node can be updated by any user that has an 'update' capability at the document-level
2. node-update:  A node can be updated by any user with a 'node-update' capability as long as they have sufficient privileges at the element-level

Element-level:

1. If a node is protected but no 'update/node-update' capabilities are explicitly granted to any user, that node can be updated by any user as long as they have 'update/node-update' capabilities at the document-level
2. If any user is explicitly granted 'update/node-update' capabilities to that node at the element level, only that specific user is allowed to update/delete that node. Other users who are expected to have that capability must be explicitly granted that permission at the element level

How does node-replace/node-delete work?

When a node-replace/node-delete is called on a specific node:

1. The user trying to update that node must have at least a 'node-update' (or 'update') capability to all the nodes up until (and including) the root node
2. None of the descendant nodes of the node being replaced/deleted can be protected by a different roles. If they are protected:
   1. 'node-delete' isn’t allowed as deleting this node would also delete the descendant node which is supposed to be protected
   2. 'node-replace' can be used to update the value (text node) of the node but replacing the node itself isn’t allowed

Note: If a caller has the 'update' capability at the document level, there is no need to do element-level permission checks since such a caller can delete or overwrite the whole document anyway.

Takeaways:

1. 'node-update' was introduced to offer finer control with ELS, in contrast to the document level 'update'
2. 'update' and 'node-update' permissions behave the same at element-level, but differently at the document-level
   1. At document-level, 'update' is more powerful as it gives the user the permission to delete the entire document
   2. All permissions talk to each other at document-level. In contrast, permissions are checked independently at the element-level
      1. At the document level, an update permission allows you to read, insert to and update the document
      2. At the element level, however, read, insert and update (node-update) are checked separately
         1. For read operations, only permissions with the read capability are checked
         2. For node update operations, only permissions with the node-update capability are checked
         3. For node insert operations, only permissions with the insert capability are checked (this is true even when compartments are used).
3. Can I use ELS without document level security (DLS)?
   1. ELS cannot be used without DLS
   2. Consider DLS the outer layer of defense, whereas ELS is the inner layer - you cannot get to the inner layer without passing through the outer layer
4. When to use DLS vs ELS?
   1. ELS offers finer control on the nodes of a document and whether to use it or not depends on your use-case. We recommend not using ELS unless it is absolutely necessary as its usage comes with serious performance implications
   2. In contrast, DLS offers better performance and works better at scale - but is not an ideal choice when you need finer control as it doesn’t allow node-level operations 
5. How does ELS performance scale with respect to different operations?
   1. Ingestion - depends on the number of protected paths
      1. During ingestion, the server inspects every node for ELS to do a hash lookup against the names of the last steps from all protected paths
      2. For every protected path that matches the hash, the server does a full test of the node against the path - the higher the number of protected paths, the higher the performance penalty
      3. While the hash lookup is very fast, the full test it comparatively much slower - and the corresponding performance penalty increases when there are a large number of nodes that match the last steps of the protected paths
         1. Consequently, we strongly recommend avoiding the use of wildcards at the leaf-level in protected paths
         2. For example: /foo/bar/* has a huge performance penalty compared to /foo/*/bar
   2. Updates - as with ingestion, ELS performance depends on the number of protected paths
   3. Query/Search - in contrast to ELS ingestion or update, ELS query performance depends on the number of query rolesets
      1. Because ELS query performance depends on the number of query rolesets, the concept of Protected PathSet was introduced in 9.0-4
      2. A Protected PathSet allows OR relationships between permissions on multiple protected paths that cover the same element
      3. Because query performance depends on the number of relevant query rolesets, it is highly recommended to use helper functions to obtain the query rolesets of nodes configured with element-level security

Further Reading

• Security Guide: Element Level Security
• Element Level Security and Redaction
• A security model for data integration

Introduction

Some customers have reported problems when attempting to access the Configuration Manager application. In the past, this has been attributed to part of the upgrade process failing for some reason (for example: a port required by MarkLogic already being used) or in some cases it was due to a default databases being removed by the customer at some previous stage.

XDMP-ARGTYPE Error

If you see this error when you attempt to access the Configuration Manager:

500 Internal Server Error XDMP-ARGTYPE XDMP-ARGTYPE: (err:XPTY0004) fn:concat( "could not initialize management plugins with scope: ", $reut:PLUGIN-SCOPE, ": ", xdmp:quote($e)) -- arg1 is not of type xs:anyAtomicType?

Resolving the error

Ensure you have an Extensions database configured by doing the following:

• Log into the MarkLogic Admin interface on port 8001 - http://[your-host]:8001/
• Under "Databases" box, ensure a database called Extensions is listed

If it does not exist, download and run the script attached to this article (create-extensions-db.xqy).

Summary

Does MarkLogic provide encryption at rest?

MarkLogic 9

MarkLogic 9 introduces the ability to encrypt 'data at rest' - data that is on media (on disk or in the cloud), as opposed to data that is being used in a process. Encryption can be applied to newly created files, configuration files, or log files. Existing data files can be encrypted by triggering a merge or re-index of the data.

For more information about using Encryption at Rest, see Encryption at Rest in the MarkLogic Security Guide.

MarkLogic 8 and Earlier releases

MarkLogic 8 does not provide support for encryption at rest for its own forests.

Memory consumption

Memory consumption patterns will be different when encryption is used:

• To access unencrypted forest data MarkLogic normally uses memory-mapped files. When files are encrypted, MarkLogic instead decrypts the entire index to anonymous memory.
• As a result, encrypted MarkLogic forests use more anonymous memory and less file-mapped memory than unencrypted forests.  
• Without encryption at rest, when available memory is low, the operating system can throw out file pages from the working set and later page them in directly from files.  But with encryption at rest, when memory is low, the operating system must write them to swap.

Using Amazon S3 Encryption For Backups

If you are hosting your data locally, would like to back up to S3 remotely, and your goal is that there cannot possibly exist unencrypted copies of your data outside your local environment, then you could backup locally and store the backups to S3 with AWS Client-Side encryption. MarkLogic does not support AWS Client-Side encryption, so this would need to be a solution outside MarkLogic.

See also: MarkLogic documentation: S3 Storage.

See also: AWS: Protecting Data Using Encryption.

Introduction

Here we compare XDBC servers and the Enhanced HTTP server in MarkLogic 8.

Details

XDBC servers are still fully supported in MarkLogic Server version 8. You can upgrade existing XDBC servers without making any changes and you can create new XDBC servers as you did in previous releases.

The Enhanced HTTP Server is an additional feature on HTTP servers which is protocol and binary transport compatible with XCC clients, as long as you use the xcc.httpcompliant=true system property.

The XCC protocol is actually just HTTP, but the details of how to handle body, headers, responses, etc., are "built in" to the XCC client libraries and the XDBC server. The HTTP server in MarkLogic 8 now shares the same low-level code and can dispatch XCC-like requests.

Introduction

This article talks about best practices for use of external proxies vs using rewriter rules in the Enhanced HTTP server.

Details

Whether to use external proxies versus using rewriter rules in the Enhanced HTTP application server is an application design tradeoff not dissimilar to using a single HTTP application server and a XQuery rewriter or endpoint that can dynamically dispatch to different databases and modules (using eval-in).  The Enhanced HTTP server does this type of dispatching much more efficiently, but the concept is similar, with the same pros and cons.

It is mostly an application and business management issue—by sharing the same port you share the same server configuration (authentication, server settings) and the "outside world" only sees one port, so configuring port-based security on firewalls, routers, or load balancers is more difficult.