em0 freezes on ZFS server

[Willem Jan Withagen]

On 26-2-2010 13:03, Jeremy Chadwick wrote:
> On Fri, Feb 26, 2010 at 10:34:41AM +0100, Willem Jan Withagen wrote:
>> This is wat netstat -m told me when it refused to revive em0:
>
> Below are the netstat -m counters/lines of concern:
>
>> 24980/2087/27067 mbufs in use (current/cache/total)
>> 24530/1070/25600/25600 mbuf clusters in use (current/cache/total/max)
>> 55305K/2801K/58106K bytes allocated to network (current/cache/total)
>
> Note how close the "current" value is to that of "total".  I'm not too
> surprised you're seeing what you are as a result of this.  What on earth
> is this machine doing at all times?
>
> Comparatively, here's some of our servers' netstat -m stats.  All these
> boxes do nightly backups to a centralised box on a private gigE network.
> All boxes use em(4).
>
> RELENG_7 amd64 2010/01/09 -- primary HTTP, pri DNS, SSH server + ZFS
>
> 	514/1931/2445 mbufs in use (current/cache/total)
> 	512/540/1052/25600 mbuf clusters in use (current/cache/total/max)
> 	1152K/6394K/7547K bytes allocated to network (current/cache/total)

That's why I wrote that I assumed that the mbuf stats where a result of the 
pipe overflowing when the device went down.
(I've seen this discussed in another thread as well)

At the moment off em0-freeze, I was running 4 bitbake compile jobs of our 
software tree. That's a lot of small file access, but like you say nothing 
really major.

--WjW