TeamServer project cannot be synchronized

I got into situation where I cannot neither update or commit project. When I click the buttons nothing happen.

Attached is my VPlog file.

Another issue. The tool failed saying something about redo history being cleaned up and that I should report the log file to you by exporting from About menu. Funny thing is that you autoexport the log file in ZIP format (which includes lot of files), but your Web UI forum doesn’t allow attach ZIP file extensions… something to be aligned.

So attached is ZIP renamed to LOG extension…, so rename or extract on linux directly
vp_15.2.zip.log (71.9 KB)

Event after restarting the application I cannot work with my project (we have hundreds of entities already in) anymore, going to delete my project folder and will try to recover from fresh install…

Hi,

I found a file was corrupted in your project folder.
Did you use teamwork files? If not, simply delete the “teamworkfiles.xml” file in your project folder.

I deleted actually yesterday whole ~/.config folder (made backup of course) and loaded project from TeamWork server as fresh copy, now 1 day later I cannot commit again and there is similar error. Cannot commit/update

The computer is running all the time without restart on Linux, latest Oracle JVM

vp.log (118.0 KB)

This time I I didn’t deleted whole .config, but tried your suggestion of teamworkfiles.xml, I did it, but after application fresh start when I do under Team -> Select repository, login to VP server nothing happen, no projects are displayed, nothing in the vp.log, its not doing anything actually…

after I exit application, I checked if teamworkfiles.xml has been even recreated, but nothing happened… its not network issue, I can access Teamserver TCP port 1999.

Things are getting really weird, I did some clean up, deleted lot of physical entities and diagram, because sync between logical and physical stopped working. Now I do mass update to TeamServer and I get error (some parse), again log attached.

We are growing quite a lot, over thousands of objects to be managed and such issues must not happen like inconsistencies… and it started :frowning:vp.log (478.3 KB)

Hi,

I found it got out of memory problem, seems your project is very large.
Please follow this article in increase memory for Visual Paradigm:

1 Like

Lol, I gave it already 4GB or RAM :slight_smile: 2 weeks ago when i noticed outofmemory…

Ok, I like such error and simple solution :slight_smile:

Thx for response.

Hello, temporarily it works, but 4GB is maximum available in UI, can you increase it to 8 or 16, etc. or is there any command line override to give more heap to VP?

Hi,

You can follow Manual approach section in above know-how.

Sorry @Kit.So - I got it…

@Kit, when I do on my linux system (Gentoo) increase of RAM, it works and in UI (About - Utilities) I can see 8GB displayed and the error disappeared.

But when we changed on MacOS (my college computer) to 8GB the value as per the document shared, in UI, we still see 4GB and the model cannot be synchronized as same parsing error. So on MacOS it still doesn’t work.

Ok, I found solution. The documentation refers to wrong file or is meant to document how to change application which is going to be installed.

ON our MacOS system I found a file called /Users/user/Library/Application Support/VisualParadigm/vplauncher.vmoptions"

I just put in this file -Xmx8192 and now it works finally. Maybe you can update your documentation.

Thx guys for helping me with our issue resolutions!

Ladislav

Hi Ladislav,

Thanks for reporting.

Hi Ladislav,

I would like to let you know the knowhow page had been updated. Once again thank you for reporting the issue.

Best regards,
Rain Wong

1 Like

Hi guys,

I noticed one thing, it looks like Linux machine (we have 2 clients) requires much higher heap space to be set up. On Max OS, we are fine with 4-8GB, on Linux as I reported I had to go over 16GB, my college with 8GB setup wasn’t able to commit his work.

When working without Teamserver(no compare, merge is happening) even on Linux its all fine.

We has suspect that this memory large requirement is bound specificaly to Linux JVM behavior. I am running build 1.8.0_202-b08 Oracle JVM.

So to work with VP I have to set on Linux heap space max to 20GB and here is what is happening:

The sync with server takes lets say 3-5 minutes. Also I noticed while having other applications closed.
Here is some picture from heapdump:

There is some high instance object generation happening in VP, I will upload heapdum from increasing moment of high ram usage to some server for you to analyze,

I cannot share heapdump with 3rd party, but attached is report from HeapHero (it will be handy to be allowed to attach also PDF files here), please rename file attached by removing .log string suffix.heapReport-2019-05-30-14-41.pdf.log (776.3 KB)

I mostly notice duplicate byte arrays deteted…

Sorry for late reply.

Seems it is same problem as:

What is the build no your college running?
Or, may he try to perform Compact Project, to check can compact project reduce the project file size? If the project file size can be reduced, I think you may ask your college to update to latest hotfix/patch build, the problem should be solved.

If your college have updated to latest build but the problem cannot be solved. May you attach his vp log again. We will check it asap.
Thanks.

FYI: In the above post, I was testing in Linux. I think the problem should not relate to Mac OS or Linux.

@peter.wong - we will try, one of my colleges installed VP just last week (so having all fixes) on MacOs, now he has set 16GB of ram and still the issue persists. Its becoming very nasty issue and I hope you will address it as I reported this back few months and it occurs on fresh installations continuously. College of mine will share here his vp.log file from MacOsx machine.

As reported during my heapdump analysis, you have duplicates of arrays in your code (a lot), I would like to hear what have you discovered in this space?