Community
Participate
Working Groups
RAStart.sh and RAStop.sh are good ideas from RAC-land that we can use. However, as RAStop.sh is currently implemented, it doesn't know about the PC, and thus the AC is killed leaving the PC orphaned. The PC needs to be stopped as well. Until the RAC=>AC transition is complete, it will need to be done "smartly", as there is no PC in RAC-land, and the same script should be made to work for both possibilities. RAStart.sh works fairly well, but I think there is something different about RAC start-up that allows the script to background the RAC in a way that doesn't work for the AC.
This is only needed when we complete new AC backwards compatibility on the Linux platform which has been deferred from 4.2.
*** Bug 139777 has been marked as a duplicate of this bug. ***
Now that we have the -shutdown switch for the new Agent Controller working on Linux we should be able to easily provide the equivalent ACStop.sh (RAStop.sh) script. While this is not required until we replace the RAC with the new AC, it is a reasonable expectation for users to find even for the new AC. There should be no need to special handle the Process Controller process as it should go away as a result of telling the AC to exit - which is believed to be working now. Need to see if RAStart.sh has issues too involved to solve quickly now with regard to starting the AC in the background. If so, we can provide ACStart/ACStop only for now and wait to resolve equivalent behavior to the RA scripts until we are ready to offer full backwards compatibility for the RAC on the Linux/IA32 platform.
this is trivial to fix. I just looked at how RAStart.sh and RAStop.sh are implemented 1. The reason that in RAStop.sh it looks like the PC is being orphaned is because that the Stop script does a ps -ae | grep RAServer which in the AC world is only a script. The parent of the PC is ACServer. 2. RAStart.sh has similiar issues where we are referencing a script and not the actually executable. 3. Another issue with Stop is that it doesn't use the shutdown feature in the AC... it does a kill -INT ... but it should work fine. The choice is copying the functionality of RAStart.sh and RAStop.sh to ACStart.sh and RAStop.sh or having the RAC scripts work for both and having Samson merge the changes into the RAC version. I would lean towards having new AC scripts
Kevin has done the work for this, but hasn't had sufficient time to test given all the other bug work. Since it is scripts and not binaries, I think this can and should go in to i4.
This must be done with BC/AC for Linux is officially supported in upcoming release. Retargetting.
*** Bug 146523 has been marked as a duplicate of this bug. ***
Note the nuance of bug 146523: Whereas when I filed this I focused on the AC leaving the PC behind, that bugzilla mentions that even if you get it to take out the PC also, it leave other agents (file transfer and TimeCollector were explicitly mentioned) behind. I agree with one bugzilla, as long as it is comprehensive in coverage and testing.
Target 4.3i2. Recall that we removed the RAStop/Start scripts from the packaging scripts in 4.2, so be sure to ping Randy when the ACStart/Stop and RAStart/Stop scripts are done.
Added "RAStart" to title to make it clear both need to be addressed. Make sure the LD_LIBRARY_PATH issues are addressed for new AC version as discussed in bug 154387.
Added symlinks from RASt{art|op}.sh to the equivalent AC versions in {releng}/master/bash/build_ia.sh.
Kevin checked in the ACStart.sh and ACStop.sh prior to Randy's work to create the symlinks, so closing this bug.
As of TPTP 4.6.0, TPTP is in maintenance mode and focusing on improving quality by resolving relevant enhancements/defects and increasing test coverage through test creation, automation, Build Verification Tests (BVTs), and expanded run-time execution. As part of the TPTP Bugzilla housecleaning process (see http://wiki.eclipse.org/Bugzilla_Housecleaning_Processes), this enhancement/defect is verified/closed by the Project Lead since this enhancement/defect has been resolved and unverified for more than 1 year and considered to be fixed. If this enhancement/defect is still unresolved and reproducible in the latest TPTP release (http://www.eclipse.org/tptp/home/downloads/), please re-open.
