-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi all,
I''m running the latest version of puppet (debian 0.22.4-2), and
I''ve
gotten to a point with my configuration where puppetmasterd needs to be
restarted every few days due to what seem like file handles not getting
closed.
I either see this set of errors:
May 22 08:08:34 localhost puppetd[2810]: Could not get config; using
cached copy
May 22 08:08:34 localhost puppetd[2810]: Could not retrieve
configuration: Uncaught exception could not open database in method
puppetmaster.getconfig
May 22 08:08:59 localhost puppetd[2810]:
(//heron/debian_base/File[/etc/apt/listchanges.conf]/source) Could not
retrieve /files/etch/apt-listchanges.conf: Too many open files -
/etc/puppet/files/etch/apt-listchanges.conf
etc...
or the following set:
May 26 08:23:50 cassowary puppetd[24866]: Could not get config; using
cached copy
May 26 08:21:47 cormorant puppetd[12952]: Could not run
Puppet::Network::Client::Master: No space left on device -
/var/log/puppet/rails.log
etc.
There is plenty of space in /var, and stopping and starting
puppetmasterd solves this.
I''m running with:
[puppetd]
factsync = true
report = true
[puppetmasterd]
storeconfigs=true
(amongst other incidentals like paths). I''ve got 18 nodes at the
moment,
and scaling up to more. I have maybe 60 files in my filestore, 14
classes and 10 components (not all of them are used).
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
iD8DBQFGWQA19n4qXRzy1ioRAoBGAJ43+2IOZzQSUyxP3DhpQKrcOsjBrACeLKRU
crPq9IcYGpydrS5VJSy8u20=IQDA
-----END PGP SIGNATURE-----