[WBEL-users] more corruption (I think)

John Morris jmorris at beau.org
Tue Jul 12 17:57:24 CDT 2005


On Sat, 2005-07-09 at 14:09, Jim Bodkikns (Dakotacom) wrote:
>    In the last 72 hours three WBEL systems have failed. All 
> installed/updated at approximately the same time. Two at customer sites and 
> one in the office. The customer systems are identical systems from PFMicro. 
> (P4's with SCSI). The system in house is a Supermicro dual Xeon with SCSI.

What exactly failed?  You aren't very clear on that.  By the time it is
setting the hostname it has already booted and made it a good way
through starting up services.  And does it lock up hard or just fail to
setup the network?  Is there actual file system corruption or just
suspected?  What does e2fsck say when run from a rescue CD?

>    Initially I assumed corruption of the filesystem given that all of these 
> systems had been in use and functioning properly for a period of time. (With 
> many reboot cycles which occur daily). And first a customer failure, 
> followed by an inhouse failure and another customer failure - within 72 
> hours.

Why so many reboots?  Unix like system should normally be run until the
kernel gets updated, the UPS goes flat, you need to change hardware,
etc.

>    The only messages seen involve the failure to set the hostname - followed 
> by boot failure, in part due to the hostname failure. (Sendmail locks for 
> example). This just suddenly occured. It is shaking my confidence in WBEL.

Something as simple as a failure in DNS can cause symptoms like you are
describing unless you mean sendmail is really locking up the machine,
not just sitting and spinning for several minutes.

-- 
John M.      http://www.beau.org/~jmorris     This post is 100% M$Free!
Geekcode 3.1:GCS C+++ UL++++$ P++ L+++ W++ w--- Y++ b++ 5+++ R tv- e* r




More information about the Whitebox-users mailing list