Prevent multiple server processes from being created by oils_ct.sh
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Evergreen |
Fix Released
|
Medium
|
Unassigned | ||
3.10 |
Fix Released
|
Undecided
|
Unassigned | ||
3.9 |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Evergreen Version: master
OS Version: Ubuntu 18.04
In the process of following the instructions on installing the SIP Server for Evergreen ILS in the documentation (over at https:/
After making the necessary changes to my settings file, rather than restarting the server using the oils_ctl "restart_sip" action, I instead (accidentaly) used the "start_sip" action on the already running server instance. This resulted in a message being displayed: "OILS SIP Server already started : PID". However, an attempt was made by the oils_ctl.sh script to start a new SIP server instance inspite of a process running an existing server instance being detected.
The new server instance failed to start on account of the existing one, however the PID associated with the existing instance was swapped out with a new one as a consequence. I noticed the changes I had made to my oils_sip.xml settings file hadn't taking affect and attempts to stop the existing server instance via `oils_ctl -a stop_sip` weren't working.
I was able to track down the PID of the existing server instance and kill it, and further research led me to the discovery of this bug.
In order to prevent this from occurring I'm suggesting we simply respect the result of the check made for running server processes before attempting to start a new server instance.
Please observe the attached patch.
tags: | added: performance |
Changed in evergreen: | |
milestone: | 3.next → none |
status: | Fix Committed → Fix Released |
Problem and fix confirmed; I've pushed the fix to master, rel_3_9 and rel_3_10. Thanks Clayton!