What are the valid values for slump hash sizes?

Document ID : KB000012453
Last Modified Date : 14/02/2018
Show Technical Document Details

1. It is possible to configure the size of hashes that are maintained by the slump process for various CA Service Desk Manager processes. 

2. The current sizes can be found in the "SLUMP_HASH" statement in the NX.env file, within each set of parentheses.

Here is a sample statement:

@NX_SLUMP_HASH=domsrvr(65536) web(8192) spelsrvr(8192) bu_daemon(131072) BOP-LOGIN(16384) kt_daemon(131072) krc_daemon(131072) keit_daemon(131072) bpvirtdb_srvr(2048) kpi_daemon(16384)

So, for example, slump processes that have a name beginning with "domsrvr" have a slump hash size of 65536.

3. As per best practices, and with consideration that increasing the size of slump hashes could increase the slump's usage of system memory, any changes should be tested and verified in a non-production environment which mirrors the production environment, before making the same changes in the production environment.




What are the valid values for the hash sizes in the SLUMP_HASH statement in the NX.env file?


CA Service Desk Manager 14.1 and below

Valid values, which are specified between "(" and ")" in the slump hash statement are:


An explanation is provided in the comments that are included in the out-of-the-box NX.env file.  The comments are as follows: 

# Size of the slump object hash.  The hash should be large enough to
# contain about twice the number of objects that will be created by
# the process.  Too small a hash size will result in a performance
# penalty and errors in the AHD log.
# The syntax for each specification is:
#    process_name_prefix(hash_size)
# where process_name_prefix is the first portion (any length) of the
# slump process-name, and hash_size is the size of the hash (must be
# a power of 2 of at least 512).  Slump uses the longest prefix that
# matches the process name.  If there is no match, it uses the
# default value of 512.

To determine a valid process_name_prefix,  run the slstat command when CA Service Desk Manager is up and running; a list of processes that are currently connected to slump would be returned. 

For example, to find a list of all processes connected to slump that have a name beginning with "web", you could run the following command:

slstat | find "web"

The results would be similar to:

web:wsp                                                           18    26 Tue Jan 03 06:55:39 2017
web:local                                                          20    28 Tue Jan 03 06:55:36 2017

When the process is started, associated messages are written to the current stdlog. 

Here is a sample message from "C:\Program Files (x86)\CA\Service Desk Manager\log\stdlog.1":

01/02 09:33:17.66 sdmpri       web-engine           7124 SIGNIFICANT  api.c                 1069 Slump object hash created with 8192 entries (2**13)

Additional Information:

Important Note: For versions 14.1 and below, the method of modifying the out-of-the-box hash sizes is to manually edit the NX.env file.  And, to ensure that the change persists, the NX.env_nt.tpl file must be manually edited with the same changes.  Thus, the first step to making any change to either of the files, is to backup the existing files so that, if necessary, the original backed up files can be restored. 

The NX.env file exists in the base installation CA Service Desk Manager directory or folder.  This base installation directory is referred to as $NX_ROOT.  The NX.env_nt.tpl file exists in the $NX_ROOT\pdmconf subdirectory.

For more information about the slump process, see the subsection titled "Message Dispatcher (sslump_nxd)" on the following page:

For more information about the conditions under which you may want to update slump hash values, review technical document TEC380512 - What is @NX_SLUMP_HASH= variable used for in NX.env?