On Nov 30, 2007, at 12:18 PM, Mark D. Nagel wrote:
> I have a client I''m testing where the puppetd process grows slowly
> (over
> a few days) to the point where it exceeds 100MB. This triggers an
> alert
> and we restart the daemon, but I''m curious whether anyone has
noticed
> similar behavior. Details are:
>
> Server: CentOS 4.5 x86 (on VMWare Server guest)
> RPM install from dlutter repo (puppet-0.23.2-1.el4)
>
> The only thing I can think of offhand is that this is related to file
> copy actions -- we copy a few large-ish directories that way (SNMP MIB
> files, for example). I''m not quite sure where to start digging to
> find
> the cause, so if anyone has any ideas, I''d appreciate it.
Can you check whether it''s buffer memory or real memory?
*nix OSes will usually buffer any of these file operations, and
there''s nothing the processes can do about it.
If it''s not buffered memory, then it sounds like a memory leak.
--
A diplomat is a man who can convince his wife she''d look stout in a
fur
coat.
---------------------------------------------------------------------
Luke Kanies | reductivelabs.com | madstop.com