K05654112: Can not generate UCS archive via CLI or GUI

Support Solution

Original Publication Date: Jan 25, 2019

Update Date: Nov 13, 2020

Details

Description

On GUI or CLI (after running the tmsh save /sys ucs command), generating UCS may hung or fail. You may see similar to below message after generating qkview within the /var/log/ltm logs:

Waiting for process config_save (pid:) to complete.

In this error message, note the following:

is the number of a process IDs (PIDs) running on the BIG-IP system.

For example:

Waiting for process config_save (pid:16672) to complete.

Environment

BIG-IP v11.5 and later

Cause

Unknown. As a result the UCS process fails to complete and may result in multiple instances of orphaned UCS save processes.

Recommended Actions

To resolve this issue, you need to end the orphaned UCS save processes. To do so, perform the following procedure:

Impact of procedure: The impact of the recommended action depends on the specific environment. F5 recommends testing any such changes during a maintenance window with consideration to the possible impact on your specific environment.

1. Log in to the command line as an administrator with root privileges with bash access. 2. Confirm that the process ID indicated is still running the config save operation:

# /var/run/config_save.pid 16672 # pstree -ap 16672 tmsh,16672 save / sys ucs lastbackup `-sh,16680 -c /usr/bin/tmsh -f save sys config partitions all >/dev/null `-(tmsh,16691)

NOTE: If the process ID indicated in the message, in the first line, is not a TMSH process, do not attempt to terminate it, it may cause the system to malfunction. Removing the recording the PID number is sufficient.

3. In the case when this is another of process with a name different from tmsh, just remove the PID file with the command in step #6 , and steps 4 and 5 are not required: # /var/run/config_save.pid

4. If the process is reported to be a tmsh process, identify any orphaned UCS save processes with the pstree command. A less verbose option makes the process IDs visible:

# pstree -p 16672 tmsh(16672)---sh(16680)---tmsh(16691)

5. Terminate all the process IDs shown in the above output by typing the following command in the bash prompt, listing all the numbers seen in parentheses:

# pid1 pid2 ... Note: pid# would be a list of PIDs shown above, in this case 16672 16680 16691

---Check the process IDs again in a short while. If necessary, end all remaining processes (if those keep running for least 30 more seconds) by running:

# pstree -p 16672 tmsh(16672)---sh(16680)---tmsh(16691) # kill -9 pid1 pid2 ...

6. Finally, remove the PID file recording the now incorrect data. # rm -f /var/run/config_save.pid

7. 7. After you successfully stop all of the orphaned UCS save processes, you can attempt to save a new UCS archive again by typing the following command syntax:

tmsh save /sys ucs

For example, to save a UCS archive file named backup.ucs, type the following command:

tmsh save /sys ucs backup

By default the archive file saves in the /var/local/ucs directory with a .ucs file extension (/var/local/ucs /backup.ucs).

8. If the issue still persists even after performing the above-mentioned steps, restarting the restjavad process may be necessary

# tmsh restart sys service restjavad

Additional Information

None.

Related Content

K41311258: Creating a UCS archive file may fail

Applies to:

Product: BIG-IP, BIG-IP LTM 15.X.X, 14.X.X, 13.X.X, 12.X.X, 11.6.X, 11.5.X