[cfarm-users] Nice on GCC202?

Jeffrey Walton noloader at gmail.com
Fri Mar 9 03:57:14 CET 2018


On Thu, Mar 8, 2018 at 6:19 PM, Baptiste Jonglez
<baptiste at bitsofnetworks.org> wrote:
> Hi,
> ...
> As far as I can remember, it's a virtual machine with only 2 CPU cores of
> the physical machine, each core having 8 "threads", although it seems to
> be quite different from Intel Hyperthreading.  Linux does not understand
> such subtleties and just sees each "thread" as a separate CPU.

This is kind of interesting in a morbid sort of way...

I've been running our test script for the last hour or so. It takes
about 4 hours to run on a modern iCore. I don't know how long it takes
on the Sparc.

Every 30 seconds or 1 minute or so the terminal ding's and this message shows:

    Message from syslogd at gcc202 at Mar  9 05:31:14 ...
     kernel:[827541.133057] watchdog: BUG: soft lockup - CPU#3 stuck
for 22s! [sshd:244056]

It looks like the issue has been reported on CentOS, Fedora, Ubuntu,
Suse, VMware. It also looks like there are a lot of culprits, from
over-commitment of resources to bad power supplies.

Jeff


More information about the cfarm-users mailing list