We are evaluating Camlistore as a CAS and have run into some concurrency 
issues. When attempting to run camput concurrently with multiple separate 
processes (this is a requirement for our usage), only one of the concurrent 
camput operations will succeed and the others will fail with the following 
error:

Could not create/open new have cache at 
/var/lib/jenkins/.cache/camlistore/camput.havecache.e62a0781661c92d42665a548ecea475335e79077.leveldb,
 
resource temporarily unavailable

We believe this is due to leveldb process-level locking. We have moved 
indexing to mongo to work around process-level locking, however, for the 
havecache and stat cache, we see no easy way to avoid process-level 
locking. We attempted disabling havecache via the debug option 
-havecache=false, 
however, this threw a different error related to stat cache:

2017/07/26 20:51:54 Could not create/open new stat cache at 
/var/lib/jenkins/.cache/camlistore/camput.statcache.e62a0781661c92d42665a548ecea475335e79077.leveldb,
 
resource temporarily unavailable

We can't find an easy way to disable stat cache to see if it addresses our 
issues.

Is there a deployment strategy for Camlistore that allows for concurrency 
at the process level?

-- 
You received this message because you are subscribed to the Google Groups 
"Camlistore" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to