LVS
lvs-users
Google
 
Web LinuxVirtualServer.org

Re: file-max/inode-max question

To: Joseph Mack <mack.joseph@xxxxxxx>
Subject: Re: file-max/inode-max question
Cc: LVS List <lvs-users@xxxxxxxxxxxxxxxxxxxxxx>
From: Derek Glidden <dglidden@xxxxxxxxxxxxxxx>
Date: Tue, 05 Sep 2000 16:56:11 -0400
Joseph Mack wrote:
> 
> Derek Glidden wrote:
> > Basically,
> > the servers really suck down file-handles and inodes, to the point that
> > we've had to increase /proc/sys/fs/file-max and /proc/sys/fs/inode-max
> > to some incredibly high values to prevent the servers from crashing with
> > errors like:
> 
> There is a squid page which tells you all the values you have to change
> to get squid running well. LVS is "somewhat" like a squid (whatever
> "somewhat" means). Are the values you need to run your LVS anything
> like the squid values?

I have no idea, but I'll go dig around the squid pages and look for that
info.  I suspect the changes are probably about the same as what I've
already done.

(Yep, assuming the same thing you're referring to are the instructions
in the FAQ about increasing "file-max" and "inode-max" via the procfs
interface.)
 
> No-one has talked about changing any of these values on this list.
> Presumably it's not a problem for most people.

That scares me.  :)  

Then either we're "the special case" for which it all just breaks,
nobody else is running LVS with lots of servers which would require
increasing file-max or inode-max, or we're pushing the limits of what
the code is capable, which makes me sad since it really isn't doing all
that much...

-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
With Microsoft products, failure is not           Derek Glidden
an option - it's a standard component.      http://3dlinux.org/
Choose your life.  Choose your            http://www.tbcpc.org/
future.  Choose Linux.              http://www.illusionary.com/


<Prev in Thread] Current Thread [Next in Thread>