occasional groove 3.1 workspace corruption/inflation

  • Thread starter Andrew van Renen
  • Start date
A

Andrew van Renen

I think it is a corruption anyway. Under some circumstances, one of the
workspaces (of the homeworking team I support) appears to lose its connection
to its replicas. At that point its symptoms are:

* complete takeover of CPU - system becomes unusable for anything else

* 'uninhibited' growth - the workspace just gets larger and larger (even
when there is no data change), and the outgoing data in Comms Manager similarly
grows

* Groove cannot be shutdown - you can eventually close the launchbar, and
you can run the exit command, but the tray job never closes down

Running GrooveClean seems to have no effect either. And once you are in
that state you are forced to use End Task to kill Groove, which probably
just seals the fate of the workspace in question.

Has this happened to anyone else? More importantly, does anyone else have
any ideas on preventing this, or recovering from it once it has happened?
 
S

Sara Xue

Hi,

First of all, I would like to make sure whether this issue only happens to
you. If other users in the workspaces still can use them properly. I would
like to suggest you can perform the following steps to resolve this issue:

Step 1:
======
Log into Groove. From the Options menu, choose Preferences, go the Account
tab, and click the Save Account As File... button to create an account file.
If you have any workspaces in which you are the only member, please Save the
workspace as an archive file. Save the files to your computer.

Step 2:
=======
Log in the administrative account of your computer. I would like to suggest
you can run the Groove clean utility first. This will remove all account
information from this device. This information will be recoverable and the
account will need to be restored from a backup or a new account will need to
be created. To run the groove clean utility, completely shut down Groove via
the icon in the system tray. If you cannot shut it from system tray, please
go to Task Manager to kill all the processes for Groove.

1. Click the Windows Start button and choose Run.
2. Browse to C:\Program Files\Groove Networks\Groove\Bin and choose
GrooveClean.exe
3. In the Run box, add "-all" to the end of the command line, so it reads:
"C:\Program Files\Groove Networks\Groove\Bin\GrooveClean.exe" -all
4. Click OK.

After the process complete, restart your computer. Then please perform next
step:

Step 3:
======
Please refer to the following article to restore the saved account and
download the data back:

Restore account:
http://docs.groove.net/htmldocs/gui...counts_account/accounts_account_restoring.htm

Downloading workspace data from another computer or workspace membe
http://docs.groove.net/htmldocs/gui...s_managing/spaces_managing_fetching_space.htm

Is this issue resolved?

Let me know the results.
 
A

Andrew van Renen

Hello Sara,

Thanks for the response. Its refreshing after the rather quiet groove.net
forums! (It is nice to see the increasing life around Groove - I do have
a soft spot for it.)

Anyway - once a workspace starts misbehaving, it affects all other copies
of that same workspace, presumably because each one is trying to absorb whatever
it is that has inflated the corrupt one. If we disable replication of that
database, the replicas are fine and remain usable.

Incidently, if you look at the telespaces directory, it is easy to identify
the problem workspace because its .xss file will be enormous (I have seen
up to 1Gb, but it seems to be dependant on available memory). The other
workspaces (with a similar data set size) will be around 100 to 200Mb in
size. It has happened to us on 3 of around 20 workspaces (each based on
more or less the same template).

Anyway - I will give your instructions a go, and hopefully I will let you
know if I can recover the latest one (the other problem workspaces have been
abandoned).

Andrew

(recovery is good, understanding why it is happenning would be wonderful!)
 
A

Andrew van Renen

In the end, I managed to do an export from the problem workspace, and I then
created a new one and imported the data up to that new one. It now appears
to be fine.

GrooveClean (without -all) on its own did not make any difference to that
workspace, though it did clear up a couple of 40k-odd 'communications' that
were just hanging around and not completing.

So workspace recovered, and I can act sooner next time so that there is less
of an impact on my users.

Still can't figure out why it happens though.

Andrew
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top