pcp
[Top] [All Lists]

Re: [announce] PCP 2.3.0-11 available for testing

To: Troy Dawson <dawson@xxxxxxxx>
Subject: Re: [announce] PCP 2.3.0-11 available for testing
From: Ken McDonell <kenmcd@xxxxxxxxxxxxxxxxx>
Date: Mon, 17 Feb 2003 14:39:33 +1100
Cc: Mark Goodwin <markgw@xxxxxxx>, <pcp@xxxxxxxxxxx>
In-reply-to: <3E4D0C87.6000809@xxxxxxxx>
Sender: pcp-bounce@xxxxxxxxxxx
Troy, ...


On Fri, 14 Feb 2003, Troy Dawson wrote:

> Howdy Ho,
> I gave this a try, and have tried both building with the src.rpm and just from
> the source.  Both build fine, and no unexpected errors, but with both ways I
> get the following error in
> /var/log/pcp/pmcd/pmcd.log
>
> -------------
> # cat pmcd.log
> Log for pmcd on thebrain.fnal.gov started Fri Feb 14 09:23:26 2003
>
> NOTICE: using /proc/partitions for disk I/O stats
> [Fri Feb 14 09:23:26] pmcd(22462) Error: Unexpected signal 11 ...

I checked your /proc/partitions and there is nothing obvious there
(same as other systems we have locally and no pmcd failures for us).
It is likely to be something else and unrelated to this NOTICE.

Is your kernel configured with modules support?

Anyway, we really need your traceback as per Mark's mail.

To make it fail, you may have to use

        pminfo -v

in another window once you've started pmcd -f in the gdb session.

> ...
> Sorry for getting in so late in the testing process, but most people here were
> happy with pcp 2.2.2 until we started doing some security scans that caused
> pmcd to chew up 100% cpu until it was killed.

Can you provide any more details on the evil security scanner?  Is this
something pmcd should be defending against, or was this a DoS scanner
on port 4321?

> ...


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