APAR status
Closed as program error.
Error description
When an agent attaches to a TEMS we have to make sure that all situations that are distributed to that agent are started on that agent. Situation distributions are stored in the Access List. In response to the on-line event from an agent, FA, a component of the TEMS, scans this list for all situations that are distributed to the agent and forwards the notifications to Sitmon, another TEMS component. If the situation has already been started, (on behalf of one or more other agents), Sitmon simply notifies the Proxy Controller, (yet another TEMS component, one that communicates directly with agents attached to the TEMS), that the situation needs to be started on that agent. This scheme works well for standard agents. But UA is a subnode application, i.e. the UA agent acts as a manager of one or more sub-applications that report directly to it. Situations can be distributed directly to subnodes instead of to the subnode manager ... and this is where the problem lies. It seems that when the subnode manager comes on-line, FA is generating Access List notifications not only for situations that are distributed directly to the manager itself, but also for situations that are distributed to subnodes of that manager. As before, if those situations are already running on the TEMS, Sitmon just taps the Proxy Controller to start it on the appropriate subnode. BUT, as it stands today, there is no guarantee that the subnode itself has actually come on-line by the time the 'start' request is sent from Sitmon to Proxy. If it hasn't, Proxy ignores the request to start the situation.
Local fix
Restart the situation or STOP/START the Universal Agent
Problem summary
The issue involves situations failing to start at subnodes of an IRA Manager. This can occur when the IRA Manager switches from one TEMS to another TEMS. When an IRA Manager switches its thrunode (switches from on TEMS to another), it is possible for situations to fail to start on the subnodes attached to the IRA Manager. This occurs when an attempt is made to distribute situations to the subnodes before registrations have been received at the HUB TEMS for the subnodes. When the IRA Manager distribution EIB log records are sent for its subnodes in response to the reporting through node change, the receiving HUB TEMS applies the distributions and attempts to start the situation that applies to the distribution. If the subnode for this IRA Manager has not yet registered at the HUB TEMS, then the situation will fail to start because the requisite metadata for the subnode is missing.
Problem conclusion
The change is to serialize metadata deployment for a switched IRA Manager to provide access list add notification on behalf of the IRA Manager subnodes only after a confirmed subnode registration has occurred at that TEMS. The fix for this APAR is included in the following maintenance vehicle: | interim fix | 6.1.0.7-TIV-ITM-IF0002 Note: Search the IBM Technical support web site for maintenance package availability
Temporary fix
Comments
APAR Information
APAR number
IZ05788
Reported component name
TEMS
Reported component ID
5724C04MS
Reported release
610
Status
CLOSED PER
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt
Submitted date
2007-10-02
Closed date
2008-08-26
Last modified date
2008-08-26
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
Fix information
Fixed component name
TEMS
Fixed component ID
5724C04MS
Applicable component levels
R610 PSY
UP
[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SSCTLMP","label":"ITM Tivoli Enterprise Mgmt Server V6"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"610","Edition":"","Line of Business":{"code":"","label":""}}]
Document Information
Modified date:
26 August 2008