A failure code is reported in the STATUS column
of the listappinstallrecs output. Some errors can
be tried again, others are terminal.
Self-describing agent installation errors that can
be tried again
The
Tivoli® Enterprise Monitoring Agents
self-describing agent service decides what types of previously failed
self-describing agent registration or installation requests are tried
again. Only the following types of failures or error codes returned
from the
Tivoli Enterprise Monitoring Server,
on behalf of a self-describing agent registration request to the monitoring
server, are tried again:
1006 |
Duplicate SDA Installation Request |
1009 |
HUB Not There |
1017 |
Temporary Installation Error |
1021 |
Server Timed Out |
Self-describing agent installation errors that can
be tried again are registration or installation requests that started,
but have not yet modified any of the Tivoli Enterprise Monitoring Server files
or internal structures.
Terminal self-describing agent installation errors
For
any error records with a STATE value of ME,
the installation is not tried again.
In
this example, the installation record for product code
11 displays
a
STATE value of
ME:
HUB/RTEMS |
PRODUCT |
VERSION |
GRPID |
ID |
IDVER |
SEEDSTATE |
STATE |
STATUS |
RTEMS_LZ |
11 |
06230000 |
5655 |
TMS |
06230000
|
Y |
ME |
1014 |
The STATE value of ME indicates
that a self-describing agent metadata installation error occurred
on the monitoring server. The monitoring server stops attempting any self-describing
agent installation for this product code until some action is taken
by the administrator to correct the error. This correction might involve IBM® Software Support. In this scenario,
you must use the tacmd deleteappinstallrecs command
to clear the self-describing agent error record after you resolve
the problem. For more information, see the IBM Tivoli Monitoring Command Reference.
To determine
whether a self-describing agent product installation failed with a
terminal error condition, run the tacmd listappinstallrecs command
by using the -e option to display error records
only. For any error records with a STATE value
of ME, the installation is not tried again.
Take
the following steps to try the self-describing agent installation
again:
- To avoid the same failure from occurring again, you must first
correct the condition that caused the installation to fail. In addition,
the monitoring server message facilities (Audit, MSG2, and RAS1 messages)
provide more information about the cause of the failure. Take corrective
action to fix the condition or contact IBM Software
Support for assistance.
- For each monitoring server, delete the
failed installation records in the application properties table by
running the tacmd deleteappinstallrecs command.
This command removes the blocking self-describing agent product installation
record. See the IBM Tivoli Monitoring Command Reference for
more information.
- When each monitoring server failed product installation record
is cleared, the monitoring server self-describing
agent facility immediately notifies any running self-describing agent
that can provide this level of product support, to try the product
installation again. For example, if the previous installation attempt
for product pc and version 06230000 failed
with a STATE of ME, and you run
the deleteappinstallrecs command, any running pc agent
for version 06230000 immediately tries the installation
again.
- Run the tacmd listappinstallrecs -t <pc> again
for product pc to determine the current installation
state.
If the self-describing agent product installation failed
at the primary hub monitoring server, the
standby hub does not attempt to install the support. Correct the reason
for the failure at the primary hub, then use the deleteappinstallrecs command
to remove the error entry from the primary hub. Clearing the error
ensures that the self-describing agent product installation is tried
again.
If the self-describing agent product installation
fails only at the standby hub, correct the reason for the failure
at the standby hub. While logged on the primary hub, you can then
use the deleteappinstallrecs command with the -n <standby_TEMS_name> option
to remove the error entry at the standby monitoring server to allow the self-describing
agent installation to try again. Deleting the application installation
record from the standby hub does not automatically trigger the self-describing
agent product installation to try again the way it does when the record
is deleted from the primary monitoring server or
a remote monitoring server. There is no problem
if the standby hub is temporarily missing the agent product support.
- If the standby hub is recycled, any missing product support is
discovered, and the self-describing agent product installation takes
place at that time.
- If the standby hub is promoted to the acting hub, self-describing
agent product installation takes place when the first agent of that
type connects to the promoted hub.
- You can force the self-describing agent installation to try again
immediately at the standby hub, by repeating the self-describing agent
installation at the primary hub. After clearing the error application
installation record at the standby monitoring server,
use the deleteappinstallrecs -a command to delete
the non-error state record from the primary monitoring server. This command repeats the self-describing
agent installation at the primary monitoring server,
triggering the standby monitoring server installation
when it completes successfully.