We are having problem to save CMS track configuration after removing a runtime which was taken away from us and is not recoverable. The CMS track config was not updated prior to this and we can no longer make track adjustments due to the old DEV runtime sys not being available. Does anyone have similar issue and know how to fix it?
error: Server icux1920.d52.lilly.com did not accept login request as admin on port 51018. Details: ERROR: Could not establish connection to server icux1920.d52.lilly.com at port 51018: Connection refused#
My understanding for removing the runtime is whatever it is available or not, CMS should remove it. Now CMS doesn't save any change made to the track.
SAP delivered a patch, we deployed it, and fixed. Now we are able to remove the runtime, and save the CMS adjustment. However the buildspace contains no DCs. This was caused by "save and restore" previous attempt to save the CMS configuration. I only see two requests in Queued state by NWDI_CMSADM, how do I get the buildspace container restored?
try the "Initialize Compartment" button at the "Build Compartment" tab of the CBS Web UI. I´m not sure that this helps but I can image it because this creates the complete buildspace from scratch.
hm.. "initialize compartment" is probably not sufficient if there are no archives at all in the build space. It will help if archives are imported (i.e. there are DCs in archive compartments) and there's some issue with building DCs in source compartments.
If there are no DCs at all in the buildspace you should probably restore the system state from CMS first.
There were two queued requests in the buildspace, after processed them, now the buildspace shows total 422 DCs, and 422 Broken DCs. It is much less than before (620 plus). All the DCs metadata show in red, half of their usage show in red too. The current track was built on NW SP8 and ESS SP5. The server was upgraded to NW SP9, we are trying to get our modification on SP5 assembled and move to new SP6 track.
Are there any DCs in the JTECHS/JEE/SAPBUILDT compartments? I think that would account for roughly 200 DCs and absence of these DCs would also explain why everything is broken.
If they are indeed missing reimport them from CMS.
Indeed there was no any DCs in the JTECHS/JEE/SAPBUILDT compartments. Last night we re-imported these three SCs. Now the track shows total 681 DCs, but thare are 106 broken DCs with error in the build log:
I've seen something similar but with SP7. I think it was some resource problem. Can you try to trigger a rebuild just for one single broken DC when CBS is not under heavy load?
I did a single broken DC rebuild a couple of times, but no luck. One customer message from SAP on this is ESS SP6 upgrade fix this issue. We upgraded the server to SP9 with ESS SP6 imported to a new track, it still has broken DCs but different error message in the build log. Now we are in the process rebuild the SP6 track buildspace container, re-import required software (SAP-JEE, SAP_JTECHES, DI_BUILDT, EP_BUILDT), I don't have the build.log information anymore. Will post it once I have.
As Pascal pointed out: use the patch for EP_BUILDT. If you have the original unpatched version you will get lots of broken DCs because of wrong ACLs.
It will not fix the xlf conversion issue, but it may solve a lot of the other broken DCs.
Is it possible to build the DC using NWDS? (In other words: is the problem limited to CBS?) If not, perhaps the Error view in the studio might show more details.
For this SP5 track, we have not imported SP9 required SCs (SAPJEE, SAPJTECHS, SAPBUILDT, EP_BUILDT). but we did imported them (SP8 version) with patches. We imported most current SP9 patches to the new SP6 track. We want to move our development on the SP5 track to assemble it, then imported to the new SP6 track based SAP XSS cookbook.
We are not able to assemble released activities with broken DCs on the SP5 track. I look a DC for ess/au/fam/cfg, Build.log still shows xlf conversion error:
Ant runtime 4.295 seconds
Ant build finished with ERRORS
The following error occurred while executing this line:
Weird, I'd definitely open an OSS message for this.
In your second log the symptom is the same, but the files are different. Since the issue is limited to CBS I would suspect resource problems, but you already wrote that a single DC build did not work either.
The only thing that I can think of to perhaps find the problem is to increase the log severity/level of the build log. To do that for a single DC you can create a "cfg" folder for that DC (in the _comp root folder) and put a file named build.properties there with a single line
loglevel=verbose
. Check in and activate. That way Ant should dump the full stacktrace of the final exception in the build log.
Regarding Force Assembly: the SCAs created by force assembly will not contain deploy or build archives for the broken DCs. If you only want to do an assembly to transport sources that may be sufficient, but I'd ask SAP for advice here.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.