Writer to keyspace {project} is corrupted due to unrecoverable problem

We are running Ubuntu 14.04.01 LTS 64bit

I completed installation yesterday and am seeing the following errors pop up at intervals on the /administration page (e.g. attached):

14:25:24: Internal error: java.lang.Error: Writer to keyspace {project} is corrupted due to unrecoverable problem

14:27:41: Internal error: Context is disposed for project: ProjectContext{ProjectId{project}}

14:28:26: Internal error: Error while importing revision MAIN-{ revision: RevisionId{project, 90acd76413829d79b05e369826166c25b19c7cd0} @ ~ 2015 Jan 30 13:34:37}-cluster=VCS_CONTENTS_CLUSTER (base: RevisionId{project, 8a16f04de0e4b1c109cf44bc99cace76085ad7f0}(Wed Feb 04 09:48:54 PST 2015), AnalyzerState{code=15, [revision info imported] [files] [constituents] }). Writer to keyspace wordfly is corrupted due to unrecoverable problem

Please advise.



Attachment(s):
error.png
7 comments
Comment actions Permalink

We are trialing a new installation and have immediately begun to see this same error with the analyser service.

Is there any follow-up on this?

(This is a linux not a windows installation)

0
Comment actions Permalink

Hi Mike,

That can be Cassandra DB failure that we still investigating. Could you please send us logs folder from Upsource home directory for the further analysis? You might post it here or send to upsource-support@jetbrains.com

0
Comment actions Permalink

Hi,

We're getting the same error message, although we're just evaluating Upsource currently. I deleted the Upsource project and repo, and recreated it, but I get the same thing.

08:56:45: Internal error: Error while importing revision MAIN-{ revision: RevisionId{cts-alt, bc3970439ea77f7c56642fa84cf45fe98016023f} @ ~ 2006 Mar 20 10:37:12}-cluster=VCS_CONTENTS_CLUSTER (base: RevisionId{cts-alt, 19131d72815a1352b99f2e60f1486905455352cc}(Wed Apr 05 15:23:56 BST 2006), AnalyzerState{code=15, [revision info imported] [files] [constituents] }). Writer to keyspace <project> is corrupted due to unrecoverable problem

When it happens, there is (I guess fairly obviously) no more revisions processed, which means we can't do further reviews, which unfortunately our evaluation is a little stuck at this point.

Is this still being investigated? Can I send you our logs?

Thanks.

Edit: This is a Windows installation using Build 1.0.12566.

0
Comment actions Permalink

For us the problem was user error.

Indexing our SVN repository caused the database to totally fill the disk.

I mounted a larger drive and restarted and it all worked magically - no errors since.

0
Comment actions Permalink

Thanks for that, much appreciated. Though unfortunately it doesn't seem to be the problem for us 240 GB free...

0
Comment actions Permalink

Hi Kieron,

Upsource currently doesn't support Windows OSs due to an issue with internal DB Apache Cassandara. According to the given log output we can say that most likely you are experiencing this issue.

Cassandra team just recently released an update that might fix this problem. It's being verified by our QA team how Upsource communicates with the new Cassandra update and for now everything goes smoothly. So new Cassandra might be bundled with next Upsource release, somewhere in May.

Till that moment I would recommend to use Linux or Mac installation.

0
Comment actions Permalink

That's great, thanks for the update.

Sadly, using a non-Windows OS is not an option available to me, so I'll keep an eye out for the new version.

0

Please sign in to leave a comment.