ADC

Upgrade considerations - SNMP configuration

The timeout parameter for an SNMP alarm is an internal option that has no impact on the alarm configuration.

Timeout parameter might appear in the SNMP alarm configurations in the running configuration (sh running) and the saved configuration (ns.conf) even if you have not made any changes to these SNMP alarm configurations.

On upgrading to a release build with the fix for the timeout setting issue, the SNMP configurations are erroneously reset to default values.

The following SNMP alarms (if configured) are impacted during an upgrade:

  • APPFW-BUFFER-OVERFLOW
  • APPFW-COOKIE
  • APPFW-CSRF-TAG
  • APPFW-DENY-URL
  • APPFW-FIELD-CONSISTENCY
  • APPFW-FIELD-FORMAT
  • APPFW-POLICY-HIT
  • APPFW-REFERER-HEADER
  • APPFW-SAFE-COMMERCE
  • APPFW-SAFE-OBJECT
  • APPFW-SQL
  • APPFW-START-URL
  • APPFW-VIOLATIONS-TYPE
  • APPFW-XML-ATTACHMENT
  • APPFW-XML-DOS
  • APPFW-XML-SCHEMA-COMPILE
  • APPFW-XML-SOAP-FAULT
  • APPFW-XML-SQL
  • APPFW-XML-VALIDATION
  • APPFW-XML-WSI
  • APPFW-XML-XSS
  • APPFW-XSS
  • CLUSTER-BACKPLANE-HB-MISSING
  • CLUSTER-NODE-HEALTH
  • CLUSTER-NODE-QUORUM
  • CLUSTER-VERSION-MISMATCH
  • COMPACT-FLASH-ERRORS
  • CONFIG-CHANGE
  • CONFIG-SAVE
  • HA-BAD-SECONDARY-STATE
  • HA-NO-HEARTBEATS
  • HA-SYNC-FAILURE
  • HA-VERSION-MISMATCH
  • HARD-DISK-DRIVE-ERRORS
  • HA-STATE-CHANGE
  • HA-STICKY-PRIMARY
  • PORT-ALLOC-FAILED
  • SYNFLOOD

These SNMP alarms configurations are impacted when you upgrade the NetScaler to the following release builds:

  • Release 11.1 build 61.2 or later
  • Release 12.0 build 61.0 or later
  • Release 12.1 build 30.1 or later
  • Release 13.0 build 51.4 or later

Example

Let’s consider an example of CLUSTER-NODE-HEALTH SNMP alarm.

CLUSTER-NODE-HEALTH SNMP alarm is set up by using the NetScaler command line:

> set snmp alarm CLUSTER-NODE-HEALTH -time 111 -state DISABLED -severity Major

> save config
<!--NeedCopy-->

This SNMP alarm configuration appears in the saved configuration file (ns.conf) as:

set snmp alarm CLUSTER-NODE-HEALTH -time 111 -state DISABLED -severity Major -timeout 86400

<!--NeedCopy-->

During an upgrade to any of the release builds mentioned above, the following error appears in ns.log file:

May 23 09:14:46 <local0.err> ns nsconfigd: __init_config_filter(): (null) line 0: No such argument [-timeout]>> set snmp alarm CLUSTER-NODE-HEALTH -time 111 -state DISABLED -severity Major -timeout 86400.
<!--NeedCopy-->

After the upgrade, the SNMP alarm configurations are reset to the default values.

Workaround

Use one of the following workarounds:

  • Before the upgrade, remove the timeout setting from the SNMP configurations in the saved configuration file (ns.conf).

  • After the upgrade, reconfigure the SNMP alarms without the timeout parameter.

Upgrade considerations - SNMP configuration