General server settings are mandatory settings that are used
to specify global policies in EKM. You can edit server settings
by editing the ekm.xml located in the EKM_HOME/conf folder.
General server settings include:
The value of ekmUrl must be set to
the URL of the EKM server. If a front-end web server is used,
the value should be set to the URL of the front-end web server.
This setting is used in certain situations. For example, when
the EKM server sends an automatic email such as those generated
during a process execution or lifecycle signoff, it uses the ekmUrl setting to create HTML links to objects that
are referenced in the emails. This setting is also used for the data
management process that manages a user’s server data and connects
it to EKM. See The Data Management Process for details.
If this setting is incorrectly specified the data management process
will fail to launch.
When the EKM server starts up, it will automatically create the default workspace specified in this setting if one does not exist. This is the workspace that all users will be signed in to by default when they sign in to EKM.
Note: If you are changing the default workspace to another existing workspace, ensure that user accounts exist for users in that workspace (see Adding and Modifying Users), or that automatic account creation is enabled for that workspace (see Automatic Creation of User Accounts). Otherwise, users will not be able to sign in to it. In this case EKM will then attempt to sign users in to another workspace in which automatic account creation is enabled, or in which user accounts exist for them.
When the EKM server starts up, it will automatically create
a user account for the user that was specified as the Designated Root
User during installation. This user will have Shared access. The Root User is the primary administrator for all workspaces
defined in EKM, and the only user who can make restricted configuration
changes.
You can change the default Root User by specifying the EKM user name of another user as defined by their OS or LDAP account. This is case-sensitive. See How User Logins Work in the EKM Installation Guide to see how EKM user names are defined.
When changing the Root User to an existing EKM user
who currently has Analyst or Basic access, that user's access level will be automatically
changed to Shared in his or her profile.
You can override the defaultRootUser setting for a specific workspace by editing the WorkspaceConfig.xml file. For details see Designating a Root User for a Workspace.
Once the Root User has been changed, and the server has been restarted, the previous EKM user will no longer be treated as the Root User, but their account will still exist.
EKM uses a bundled content management system for managing
files and metadata. The repository setting
in ekm.xml can be used to configure the retentionPolicy setting.
The retentionPolicy setting specifies
the global default expiration time (in days) for those objects that
are marked as perishable in EKM. It can
also be used to specify the default expiration time on a per-type
basis. Figure 6.5: Default Values for retentionPolicy Setting shows
the default values for this setting. You will seldom need to change
this value.
In the listing, the default expiration time for any perishable
object is 365 days. For objects of type File or Folder, the expiration time is 1825 days (or 5
years). For an Audit Log, the
default expiration time is 30 days. For a Job object it is 30 days, and for a Data Extraction Monitor object it is 7 days. Jobs are not marked
as perishable until they have finished executing.
Figure 6.5: Default Values for retentionPolicy Setting
<repository>
<retentionPolicy>
<defaultExpiration>365</defaultExpiration>
<customRetentionSetting>
<objectType>AuditLog</objectType>
<expiration>30</expiration>
</customRetentionSetting>
<customRetentionSetting>
<objectType>File</objectType>
<expiration>1825</expiration>
</customRetentionSetting>
<customRetentionSetting>
<objectType>Folder</objectType>
<expiration>1825</expiration>
</customRetentionSetting>
<customRetentionSetting>
<objectType>Job</objectType>
<expiration>30</expiration>
</customRetentionSetting>
<customRetentionSetting>
<objectType>DataExtractionMonitor</objectType>
<expiration>7</expiration>
</customRetentionSetting>
</retentionPolicy>
</repository>
Note: The retention setting for Job objects does not
apply to jobs submitted to EKM from Workbench. Such jobs expire when they are
released by RSM, as detected by the scheduled batchJobMonitor task.
For more information about how these types of jobs are handled, see Submitting Jobs to EKM from Workbench in the EKM User’s Guide.
The modules setting configures modules
that provide groupings of functionality. The default value for this
setting is shown below.
Figure 6.6: Default Value for modules Setting
<modules>
<module name="ekmServer" enabled="true"/>
</modules>When defining a module you need to specify a module
name, and specify whether or not the module is enabled
by entering either a true or false value. ekmServer is the only module that
is defined, and it should be enabled.
The localprocess setting specifies policies
for executing a predefined set of external applications directly on
the EKM server that perform metadata and report extraction on
simulation files. The default values for this setting are shown below.
Figure 6.7: Default Values for localProcess Setting
<localProcess>
<dataExtractionTimeout>3600</dataExtractionTimeout>
<dataExtractionThreads>1</dataExtractionThreads>
<maxDataExtractionMemory>16</maxDataExtractionMemory>
</localProcess>
dataExtractionTimeoutSpecifies the timeout value (in seconds) for the predefined set of metadata and report extraction applications. If an application takes longer than the timeout value to complete, it will be terminated by EKM. The default value is 3600 seconds (or 1 hour).
dataExtractionThreadsThe maximum number of simultaneous data extractions
that can occur on the EKM server (or cluster node in the case
of a cluster setup). To prevent data extractions from occurring, set
this value to 0.
maxDataExtractionMemoryThe maximum amount of memory (in GBs) that can be
used for data extraction. This is the cumulative memory of all extraction
threads. The value should be based on the available RAM as well as
the memory required by EKM and other processes on the server.
For example, for a 64GB server, this value might be set to 48 to keep the server functioning optimally when data
extraction is taking place.
If the memory required to
extract data from a file exceeds the maxDataExtractionMemory limit, data will not be extracted from that file. For Fluent (.cas) and CFX (.def, .res) files, which support partial data extraction, metadata may be extracted
from the file but not a report or image if there is insufficient memory.
The same applies to the component files of a Workbench archive (.wbpz) file.
The remoteProcess setting specifies
policies for executing external applications on a remote machine using
ANSYS' Remote Solve Manager (RSM). The default values for this setting
are shown below.
Figure 6.8: Default Values for remoteProcess Setting
<remoteProcess>
<jobSourceRootPath>{$EKM_TEMP}/jobRoot</jobSourceRootPath>
<!-- Uncomment the following to specify the default starting folder for
job file system as User’s Home folder (UserHome) instead of default user’s jobRoot (JobRoot) -->
<!-- <jobFileSystemHome>UserHome</jobFileSystemHome> -->
<clusterVisibleDirectories>
</clusterVisibleDirectories>
<rsmWindowsDomain>MYDOMAIN</rsmWindowsDomain>
<rsmLauncherURL>http://localhost:10182/LauncherService</rsmLauncherURL>
<jobOutputFilename>ekm_job.out</jobOutputFilename>
<dataManagementJreLocation>%AWP_ROOT182%\EKM\programs\jre1.8.0_121</dataManagementJreLocation>
<dataManagementJarLocation>%AWP_ROOT182%\EKM\bin\ekm-servo-18.2.jar</dataManagementJarLocation>
<!--
Uncomment the following to integrate with EnginFrame for remote visualization.
You will need to change the following:
a) In serviceDefinitionFile, set the fully qualified host name of the server where
enginframe is running
b) Change the name of the adminUser if its different from efadmin
c) Change the queue definitions. The template provides definition for two queues:
a windows queue using NICE Neutro as the scheduler and NICE DCV as remote visualization solution;
and a linux queue using Platform LSF (or openlava) as the scheduler and NICE DCV as remote
visualization solution. You may need to modify, add or remove queue definitions as needed for
your configuration.
-->
<!--
<enginframe>
<serviceDefinitionFile>
https://<enginframe-server-name>:8443/enginframe/ekminteractive/ekminteractive.xml
</serviceDefinitionFile>
<adminUser>efadmin</adminUser>
<queue>
<name>Windows Remote Viz Queue</name>
<os>windows</os>
<scheduler>neutro</scheduler>
<schedulerQueue></schedulerQueue>
<remoteType>dcv</remoteType>
</queue>
<queue>
<name>Linux Remote Viz Queue</name>
<os>linux</os>
<scheduler>lsf</scheduler>
<schedulerQueue></schedulerQueue>
<remoteType>dcv</remoteType>
</queue>
</enginframe>
-->
<workbenchServerPortRange>40000-40100</workbenchServerPortRange>
<!-- Uncomment the following to show the CycleCloud cluster monitor under the
Cluster Monitors section under Jobs page of EKM UI -->
<!--
<clusterMonitors>
<clusterMonitor>
<type>CycleCloud</type>
<param name="url" value="http://<cycle-cloud-server-name>/node_status"/>
<param name="credentialInfoFilePath" value="{$EKM_HOME}/conf/cycleCloudCredentials.json"/>
<param name="clusterName" value="kumo-hpc"/>
<param name="hostPrefix" value="ip-"/>
<param name="excludeNodeArrays" value="master,sgemaster,app,execute"/>
</clusterMonitor>
</clusterMonitors>
-->
<!--
The following solver versions correspond to the installed versions of ANSYS solvers for submitting jobs using built-in templates.
To add support for more versions add an extra <version> element with <value> and <labels>
For example, to add version support for 16.1, add the following inside <solverVersions>
<version>
<value>161</value>
<label>16.1</label>
</version>
The default value for the version can be specified as an attribute of the solverVersions element.
-->
<solverVersions default="182">
<version>
<value>182</value>
<label>18.2</label>
</version>
<version>
<value>180</value>
<label>18.0</label>
</version>
<version>
<value>172</value>
<label>17.2</label>
</version>
</solverVersions>
</remoteProcess>
jobSourceRootPathThe path to the directory where EKM will stage the data files for batch jobs submitted to RSM, if users do not choose a working directory when setting up jobs. This directory is specified in the EKM job data directory field during installation, and is resolved on the cluster submit host, so it must be accessible from that machine. See The Creation and Structure of Job Directories for details.
In Windows:
If EKM and RSM are running on Windows and integrating with a Windows HPC cluster, this
path must be a UNC path to a shared directory (for example,
\\HEADNODE\HPCTemp). All cluster nodes should be
able to access this share. In this case, EKM will create a
workspace directory under the job data
directory, and username directories will be
created for each user under that directory (for example,
\\HEADNODE\HPCTemp\Default\jsmith).
In Linux:
If EKM and RSM are running on Linux and integrating with a Linux cluster, this can be an
absolute path to a shared directory, or a path that is relative to
users’ home directories. The use of home directories (via a relative
path) is strongly recommended. A user’s home directory is determined
by the cluster submit host, and the user has read/write permission on this
directory and any subdirectories that it contains. For example, specifying
the value jobRoot would translate into
/path/to/user/home/jobRoot for each
user.
In either Windows or Linux, if you specify an absolute path to a shared job data directory, you must make considerations regarding access and security, as described in Access and Permission Requirements of the Job Data Directory.
You must create the job data directory specified in
the jobSourceRootPath setting before job
submission begins. EKM will not create it for you.
If you are using a shared directory in either Windows
or Linux, it is recommended that you create the workspace directory up front, and ensure that all users have read/write permission
on that directory, so that username directories
can be created under it. Additional steps may also be required depending
on your security requirements. For more information, see The Creation and Structure of Job Directories and Access and Permission Requirements of the Job Data Directory.
The EKM server, RSM launcher service and cluster software must all be running on the same O.S. For example, a Linux cluster will not work with an install of RSM and EKM on Windows.
If users will be submitting jobs to EKM from Workbench,
and Workbench and RSM are on different machines that refer to
the same physical job directory using different paths, then you
must specify a pathMapping in the
ekm.xml file. This ensures that
EKM passes a job path to RSM that RSM can recognize. For
more information see the pathMappings
setting in this section.
jobFileSystemHomeThis setting is optional. It enables you to change the default starting directory on the > page.
If included, there are two options:
Set the value to UserHome to set the
default starting directory on the File System
page to the user's home folder (for example,
/nfs/testMachine/userData/user1 or
C:\Users\user1).
<jobFileSystemHome>UserHome</jobFileSystemHome>
Set the value to JobRoot to set the
starting directory on the File System page to
the EKM job data directory (in other words, the directory specified
in the jobSourceRootPath setting).
<jobFileSystemHome>JobRoot</jobFileSystemHome>
For information about the File System page, see Accessing the File System.
clusterVisibleDirectoriesYou only need to specify this setting if users will be choosing working
directories for batch jobs (instead of using the auto-created directories),
or if jobSourceRootPath is set to a relative path
for a Linux installation.
If jobSourceRootPath is a full/absolute path,
EKM automatically considers the path to be visible, so you do not need
to specify that path in the
clusterVisibleDirectories setting.
EKM assumes that it is submitting jobs to either an ANSYS RSM Cluster
(ARC) or third-party cluster. Use the
clusterVisibleDirectories setting to list
directories that are visible to all execution nodes. This determines where
the job data directory can reside, enables a job to run directly in its
specified working directory, and prevents unnecessary file transfers to/from
RSM.
In the example below, users can run jobs in paths under
/net/userhomes,
/spdm/wip, and /home.
These are directories that all cluster nodes can access. The
/net/userhomes directory, for example, is a
shared directory under which all users’ home directories are located.
Specifying this directory as a visible cluster directory means that
directories such as /net/userhomes/jsmith and
/net/userhomes/mjones are also visible and
valid. Job data for these users will be stored in
/net/userhomes/jsmith/jobdata and
/net/userhomes/mjones/jobdata.
You can specify as many paths as needed in the
clusterVisibleDirectories setting. When adding
paths, make sure that you enclose each path in a
<dir> tag, as shown below.
<jobSourceRootPath>jobdata</jobSourceRootPath>
<clusterVisibleDirectories>
<dir>/net/userhomes</dir>
<dir>/spdm/wip</dir>
<dir>/home</dir>
</clusterVisibleDirectories>If EKM and RSM are integrating with a Linux cluster, you can specify
an absolute or relative path for jobSourceRootPath.
You must always specify full paths for
clusterVisibleDirectories.
pathMappingsThis is an optional setting that is used if users submit jobs to EKM
from Workbench. You will need to add this setting if Workbench and RSM are
installed on different machines, and use different paths to refer to the
same physical job directory. The pathMappings
setting enables you to map directory paths on the Workbench machine to paths
used on the RSM machine or cluster node where jobs run. Each mapping
consists of a pair of settings: the clientPath and
clusterPath, as shown in the example
below.
A user submits a job from Workbench to EKM. Workbench refers to the job’s directory as follows:
\\clusterServer1\jobsShare\Job_15 Jun 2017 - 09.56.59
AM
EKM must be able to see this path.
On the cluster node where RSM will eventually run the job, RSM refers to the same directory in a different way:
\\clusterServer1\jobs\Job_15 Jun 2017 - 09.56.59
AM
The pathMapping for this setup would
consist of the following pair of settings:
<pathMappings>
<clientPath>\\clusterServer1\jobsShare</clientPath>
<clusterPath>\\clusterServer1\jobs</clusterPath>
</pathMappings>
With this pathMapping, EKM will
convert the Workbench path (clientPath) to
the RSM version of the path (clusterPath)
before it passes the job to RSM.
In this case, this means that EKM will convert
\\clusterServer1\jobsShare\Job_15 Jun 2017 -
09.56.59 AM to
\\clusterServer1\jobs\Job_15 Jun 2017 - 09.56.59
AM, and pass the latter path to RSM.
Important: Any clusterPath that you specify must be an
allowed path specified in the
clusterVisibleDirectories setting.
rsmWindowsDomain (Windows only)Use this setting to ensure successful job submission to RSM when RSM is running on Windows.
This is the Windows domain name that will be prepended to the user name of
all EKM users when batch jobs are submitted to RSM. All user jobs will
be submitted to RSM as rsmWindowsDomain
value\OS account name. For
example, if MYDOMAIN is specified as the value for
rsmWindowsDomain, and a user’s OS
account name is jsmith, then jobs will be submitted
to RSM as MYDOMAIN\jsmith.
Note: It is recommended that you use a NetBIOS domain name rather than a Fully Qualified Domain Name (FQDN) for this setting.
jobTemplateUse this optional setting to change the default RSM job template file used to run EKM jobs to a customized one.
rsmLauncherURLIf the cluster submit host is not the EKM server, you must change the value of this
setting from localhost to the host name of the
cluster submit host (where the ANSYS RSM Launcher service is
running).
You must also change the value of this setting if the default port of the RSM launcher service is changed (see Specifying a Port for the Launcher XmlRpc Listener (EKM-RSM) in the RSM User's Guide).
Example:
The RSM launcher service configuration is modified to use port 10099
instead of the default 10182. The
Ans.Rsm.AppSettings.config file in the
directory is edited as follows:install_root\RSM\Config
</appSettings>
<appSettings name="LauncherService">
<add key="LauncherXmlRpcEnabled" value="true" />
<!-- Port for Launcher XmlRpc listener. http://+:port/Launcher -->
<add key="LauncherXmlRpcPort" value="10099" />
<add key="LauncherXmlRpcSecure" value="false" />
<add key="XmlRpcMaxConcurrentRequests" value="10" />
<!--
Range of ports for user XmlRpc listeners. http://+:port/UserProxy
-->
<add key="ProxyXmlRpcPortRange" value="50000-50100" />
<add key="ProxyXmlRpcLogMessagePrefix" value="[P] " />
</appSettings>The rsmLauncherURL setting in ekm.xml is also
modified to specify 10099 as the port in the
URL:
<rsmLauncherURL>http://localhost:10099/rsm/</rsmLauncherURL>rsmClientTimeoutSpecifies how long in seconds EKM should wait for a response from the RSM launcher service (see the rsmLauncherURL setting) before giving up and reporting an error.
This setting is optional. If included, the
value must be positive. A value of 0 will cause EKM to
wait for a response indefinitely. If this setting is not included,
the timeout used will be 120 seconds.
Example:
<rsmClientTimeout>90</rsmClientTimeout>
jobOutputFilenameThe file name that will be used by all jobs to capture batch processing output.
dataManagementJreLocationThe full path to the Java Runtime Environment (JRE) used to run the data management process. For details see The Data Management Process.
<dataManagementJreLocation>%AWP_ROOT182%\EKM\programs\jre1.8.0_121</dataManagementJreLocation>
By default this path is set to the full path of the JRE (JRE Home) included under the EKM folder under the ANSYS installation root folder. In general, you should not need to change this setting.
The AWP_ROOT182 variable in the path will be resolved on
the cluster submit host (where the RSM launcher service is running).
Example:
AWP_ROOT182 = C:\Program Files\ANSYS
Inc\v182
RSM will use a JRE located at C:\Program Files\ANSYS Inc\v182\EKM\programs\jre1.8.0_121.
The dataManagementJreLocation should always be set to a valid local
path on the cluster submit host, as the Data Management process
is launched on that machine.
You do not need to change the path format of the dataManagementJreLocation for a Linux installation.
RSM will convert the Windows path into a Linux path automatically.
dataManagementJarLocationThe full path to the executable JAR file used to run the data management process.
<dataManagementJarLocation>%AWP_ROOT182%\EKM\bin\ekm-servo-18.2.jar</dataManagementJarLocation>
By default this path is set to the full path of the ekm-servo-18.2.jar file included under the EKM folder under the ANSYS installation root folder. In general, you should not need to change this setting.
The AWP_ROOT182 variable in the path will be resolved on
the cluster submit host.
Example:
AWP_ROOT182 = C:\Program Files\ANSYS
Inc\v182
RSM will use a JAR located at C:\Program Files\ANSYS Inc\v182\EKM\bin\ekm-servo-18.2.jar.
The dataManagementJarLocation should always be set to a valid local
path on the cluster submit host, as the Data Management process
is launched on that machine.
You do not need to change the path format of the dataManagementJreLocation and dataManagementJarLocation for a Linux installation. RSM will convert the Windows path into
a Linux path automatically.
enginframeContains EnginFrame settings and interactive queue definitions. For more detailed information see Configuring EKM to Work with EnginFrame in the EKM Installation Guide.
Figure 6.9: Sample Interactive Queue Definitions
<enginframe>
<serviceDefinitionFile>https://
enginframe-server-name:8443/enginframe/ekminteractive/ekminteractive.xml
</serviceDefinitionFile>
<adminUser>efadmin</adminUser>
<queue>
<name>Windows Remote Viz Queue</name>
<os>windows</os>
<scheduler>neutro</scheduler>
<schedulerQueue></schedulerQueue>
<remoteType>dcv</remoteType>
</queue>
<queue>
<name>Linux Remote Viz Queue</name>
<os>linux</os>
<scheduler>lsf</scheduler>
<schedulerQueue></schedulerQueue>
<remoteType>dcv</remoteType>
</queue>
</enginframe>The interactive queues defined in the ekm.xml file will be automatically created in the /Administration/Servers/Master folder when the EKM server is started. Every interactive queue
that is created in EKM will automatically contain the predefined
application remoteDesktop.app.xml, which
launches an empty remote desktop session. Once interactive queues
have been created in EKM you can define job submission settings for
each queue, including the account names that will be used to launch
remote sessions.
workbenchServerPortRangeThe reserved range of ports that the Workbench server will use for listening to command requests from EKM. This setting is used when Workbench server jobs are started from EKM. The range is defined using a min-max format, as shown in the example below:
<workbenchServerPortRange>51215-51220</workbenchServerPortRange>
clusterMonitorsWhen EKM is deployed on the Cloud, configuring CycleCloud in the ekm.xml file enables EKM users to monitor the HPC cluster. When configured correctly, a Cluster Monitor page is available in the Jobs section in EKM, and cluster monitor graphs are displayed on the HPC tab on that page. See Cluster Monitoring in the EKM User’s Guide for details.
<clusterMonitors>
<clusterMonitor>
<type>CycleCloud</type>
<param name="url" value="http://<cycle-cloud-server-name>/node_status"/>
<param name="credentialInfoFilePath" value="{$EKM_HOME}/conf/cycleCloudCredentials.json"/>
<param name="clusterName" value="kumo-hpc"/>
<param name="hostPrefix" value="ip-"/>
<param name="excludeNodeArrays" value="master,sgemaster,app,execute"/>
</clusterMonitor>
</clusterMonitors>The <type> should remain set to CycleCloud.
Specify the parameters described below:
url: The URL of the REST
API endpoint that will be queried for data
credentialInfoFilePath: The
path to the .json file where encrypted credentials
will be stored
clusterName: The name of
the cluster to be monitored
hostPrefix: The prefix appended
to host names
excludeNodeArrays: A comma-separated
list of node arrays that you do not want to be included in the Cluster Monitor
(optional) maxHttpClientConnections: The
maximum number of concurrent connections between the HttpClient and
CycleCloud
The HttpClient retrieves data from CycleCloud for the purpose of
cluster monitoring. By default, the maximum number of concurrent
connections is 10. If after some time you find that the number of
user visits to the Cluster Monitor page is
increasing, and 10 concurrent connections is not enough to handle
the load, it is recommended that you add the
maxHttpClientConnections setting and
increase its value to allow more concurrent connections. Note that
if several requests are made in quick succession, new connections
will not be created for each one. Rather, connections will be reused
when possible.
<param name="maxHttpClientConnections" value="15"/>
solverVersionsThe solver versions specified here correspond to the installed versions of ANSYS solvers and applications that are used when jobs are submitted using EKM’s built-in CFX Batch, Electronics Batch, Fluent Batch, LS-DYNA Batch, MAPDL Batch, and Workbench Batch job templates.
<solverVersions default="180">
<version>
<value>182</value>
<label>18.2</label>
</version>
<version>
<value>172</value>
<label>17.2</label>
</version>
</solverVersions>
Solver versions specified in <version> elements will appear in the
Version drop-down list in the Start
Job dialog box when a built-in job template is used to start
a job. The default attribute determines which
version will be selected by default in this list.

To add support for more solver versions, create additional
<version> elements with
<value> and
<label> attributes. For example, to add
support for version 17.0 solvers, you would add the following
<version> element:
<version>
<value>170</value>
<label>17.0</label>
</version>
The <value> is what EKM uses to refer to the version
internally, while the <label> is what is
displayed in the Version drop-down list in the
Start Job dialog box.
Make sure that the versions specified correspond to the actual ANSYS application versions installed on your execution nodes.
The <solverVersions> element
is intended for ANSYS products only. Non-ANSYS solvers are not supported.
For on-premise installations, versions within the past two major releases are supported (in other words, release 16 and 17). For Cloud installations, versions 16.1 and higher are supported.
The Workbench Batch template does not
support job submission to versions of Workbench older than
version 17.0. Pre-R17 versions will not be available in the
Version drop box when using this
template, even if such versions are specified in the
<solverVersions> setting.
enableClusterJobNameSpecificationAdding this setting to the <remoteProcess>
settings allows a custom field to be displayed in the job submission wizard
which ties the EKM job to a specified account name that can be used for
tracking cluster usage.
<enableClusterJobNameSpecification>true </enableClusterJobNameSpecification>
For information on defining a cluster job name field and tracking cluster usage, see Tracking Cluster Usage Through a Cluster Job Name.
The cache setting specifies policies
for storing files cached from remote EKM servers. The default
value for this setting is shown below.
sizeThe maximum amount of data, in bytes, that can be cached. You can use suffixes k or K for kilobytes, m or M or megabytes and g or G for gigabytes. For example, 20G shown in the above example denotes 20 gigabytes.
Note: A cache size of 0 means that
no files will be cached.
The scheduledTasks setting specifies
the day and time when periodic tasks are executed by EKM. Below
are the default values for this setting, followed by a description
of each setting.
Figure 6.11: Default Values for scheduledTasks Setting
<dataStoreGc>
<day>Sunday</day>
<hour>0</hour>
<dayInterval>1</dayInterval>
</dataStoreGc>
<fileBackup>
<day>Sunday</day>
<hour>0</hour>
<dayInterval>0</dayInterval>
</fileBackup>
<remoteSync>
<day>Any</day>
<hour>1</hour>
<dayInterval>1</dayInterval>
</remoteSync>
<recycleBinCleanup>
<day>Any</day>
<hour>0</hour>
<dayInterval>1</dayInterval>
</recycleBinCleanup>
<mail>
<day>Any</day>
<hour>20</hour>
<dayInterval>1</dayInterval>
</mail>
<purge>
<day>Saturday</day>
<hour>22</hour>
<dayInterval>7</dayInterval>
</purge>
<batchJobMonitor>
<day>Any</day>
<hour>0</hour>
<dayInterval>0</dayInterval>
<minInterval>1</minInterval>
</batchJobMonitor>
<serverMonitor>
<day>Any</day>
<hour>0</hour>
<dayInterval>0</dayInterval>
<minInterval>10</minInterval>
</serverMonitor>
<cacheMonitor>
<day>Any</day>
<hour>23</hour>
<dayInterval>1</dayInterval>
</cacheMonitor>
</scheduledTasks>
You can specify the execution time for any task by providing
the day, hour, and min values. Any value can be
used for day if you want the task to be executed
all days. Alternatively, you can specify a particular day of the week
such as Saturday, Sunday, and so on. The hour values range from 0 (12 AM) to 23 (11 PM). The min (minute) value ranges from 0 to 59. If min is
unspecified, it is taken as 0. You can specify
the interval of the task by providing the dayInterval and hourInterval values. If you set both
of these values to 0, the task will never be
executed.
dataStoreGCUsed to periodically clean files from the datastore folder that are no longer being used within EKM. When a file is deleted from EKM it is not removed immediately but is marked for garbage collection. The file content is actually removed as part of this scheduled task. As shown in Figure 6.11: Default Values for scheduledTasks Setting, this task is executed every day at 2 AM.
To ensure consistency during restore and preventing accidental deletion of files that have not yet been committed, this task will delete only those files that are no longer in use, and whose modification time is older than garbage collection start time minus a buffer time. The buffer time is equal to the specified interval of dataStoreGC scheduled task plus an additional buffer of 6 hours. For example, suppose default settings are used for dataStoreGC and a file was added on Monday at 3PM. The next garbage collection will start on Tuesday at 2AM, but this file will not be deleted because it is still in use. Now suppose this file is deleted on Tuesday at 10AM. The next garbage collection will start on Wednesday at 2AM, but because the file’s modification time is less than the buffer of 30 hours (24 hour dataStoreGC interval and 6 hour of additional buffer), it will not be deleted even then. The file will only be deleted in the next garbage collection on Thursday at 2AM. This ensures that if a complete backup was started before 10 AM on Thursday, the file will exist in both the database backup as well as the datastore backup, as long as the backup finishes before Thursday at 2AM.
Important: The garbage collection interval should exceed the time it takes to complete a full backup of the datastore to avoid data inconsistency during restore. Thus if the backup time exceeds 24 hours, the dayInterval setting should be increased from the default value of 1.
fileBackupUsed to create a backup of all files uploaded to EKM.
This is desirable if you want to periodically write all files managed
by EKM to a separate, backup folder. Because you may decide to
use your own backup policy, by default, this task is specified to
never execute because the dayInterval and hourInterval values are set to 0.
remoteSyncUsed to periodically synchronize all files and folders in remote file servers. EKM allows users to create a reference to files/folders that reside in remote servers, in lieu of uploading them. This task is used to periodically poll the external servers and synchronize the information in EKM, to determine if any changes have occurred in the file server.
Important: Do not schedule the remoteSync task to run at midnight. Doing so can interfere with the creation
of the daily audit logs.
recycleBinCleanup Used to periodically purge objects from the Recycle Bin of all users in all workspaces, based on the recycle bin cleanup policy specified in the users' preference. Objects will be deleted only if the user has chosen to delete objects in the Recycle Bin that are older than X days, and more than X days have elapsed because the object was placed in the Recycle Bin.
mailUsed to send an email to users that have subscribed to receive alerts for certain objects. By default, this task runs at 8 PM every day.
purgeUsed to remove objects from the repository that have expired. By default, this task runs every Saturday at 10 PM.
batchJobMonitorUsed for monitoring the status of asynchronous batch jobs submitted to RSM. This task is run every minute to get real time feedback.
serverMonitorUsed for monitoring the status of remote EKM servers. This task is run every 10 minutes. If a server cannot be reached, a warning is shown in the EKM user interface, alerting the system administrator to investigate the issue.
cacheMonitorUsed to monitor the size of the cached data. It is run every night at 11 PM. It first removes any cached files that are no longer present in the master EKM workspace where the file came from. It then calculates the total size of the cache and if the size exceeds the maximum limit, it removes the oldest files one at a time until the size is below the limit. The “age” of the file is based on its last access time.
The pluginRegistry setting is used
to configure plug-ins that are loaded by EKM. The default values
for this setting are shown below.
Figure 6.12: Default Values for pluginRegistry Setting
<pluginRegistry>
<pluginDirectory>
{$EKM_HOME}/plugins-deploy
</pluginDirectory>
<loadAll>true</loadAll>
<pluginSettings pluginName="cae">
<param name="customResourcesPath" value=""/>
</pluginSettings>
</pluginRegistry>
pluginDirectorySpecifies the full path of the directory containing
the plug-in JAR files. By default, it is set to the plugins-deploy directory within the EKM installation directory. The loadAll setting determines whether all the plug-in JAR
files in the pluginDirectory should be loaded
or not. By default, this value is set to true. If you specify it as
false, then you can explicitly list all the plug-ins you want to load
by the following line. You will seldom need to change these two values.
<plugin>PLUGIN-JAR-NAME</plugin>
pluginSettingsUsed to specify plug-in-specific settings that are
given as name value pairs. Currently the
only available setting is customResourcesPath for the cae plug-in. The customResourcesPath setting is used to specify a path where script files for metadata
extraction and report generation can be placed if you want to override
the default data extraction of built-in CAE types.
The superUser setting is used to specify
the user name and password of the Superuser. The Superuser has access
to the EKM Administration interface, which
allows workspace creation and migration.
Figure 6.13: Default Value for superUser Setting
<superUser>
<name>superuser</name>
<password>9a85a0cc0a56febcf46690a6ab1a5803</password>
</superUser>
nameThe name of the user who is the designated Superuser. This setting is optional. Only a password is required to sign in to the EKM Administration interface.
passwordSpecifies the md5 hash of the superuser password.
For security reasons clear text password is not stored in the file.
The default value corresponds to the password: superuser123. It is recommended that you use the administration interface to
change this default password after installation.
For information on the EKM Administration interface, see Creating and Managing Workspaces.
The cluster setting is used to specify
the configuration of a cluster.
Figure 6.14: Default Value for cluster Setting
<cluster>
<type>cluster</type>
<cacheConfigFile>{$EKM_HOME}/conf/clusterCache.xml</cacheConfigFile>
</cluster>
typeMust be cluster.
cacheConfigFileSpecifies the location of the configuration file for the infinispan cache, which is a cluster-wide in-memory cache.
The maxSearchResults value in the memoryLimits setting is used to limit the number of
objects that are returned by a search. This limits the amount of time
and resources that are spent on search operations. If your server
has a large amount of memory available and you want to allow users
to search for more objects, you can increase the limit. The default
value for this setting is shown below.
Figure 6.15: Default Value for maxSearchResults Setting
<memoryLimits>
<maxSearchResults>500</maxSearchResults>
</memoryLimits>
The email setting enables you to specify
a configured domain for all e-mails, alerts and messages that are
sent by EKM. By default, the domain that you specify will be
part of every e-mail address specified in EKM. The default value
for this setting is shown below.
Figure 6.16: Default Value for email Setting
<email>
<mailFromDomain>nodomain.com</mailFromDomain>
</email>The customResourcesPath setting is
used to specify the folder that contains user interface files for
custom dialog boxes. Refer to the Defining Custom Dialogs, Wizards and Applications chapter in the EKM User’s Guide for details on how to create these files.
The default path shown is a suggested location for the custom resources folder. You can either create a folder in this location, or specify a path to a different folder on the EKM server.
Figure 6.17: Default Value for customResourcesPath Setting
<customResourcesPath>{$EKM_HOME}/conf/resources</customResourcesPath>The fileTransferService setting determines
the root directory for the file transfer service when uploading a
file to EKM using the File Transfer Client. This is the staging
area in which files are temporarily held before being imported into EKM.
The default value is the TEMP directory.
In a clustered environment, the rootDir value
will need to be changed on each node to a shared directory that all
nodes can access.
Figure 6.18: Default Value for fileTransferService Setting
<fileTransferService>
<rootDir>{$EKM_TEMP}</rootDir>
</fileTransferService>