6 Sep
2011
6 Sep
'11
10:22 p.m.
Hi Mathew,
On Tue, Sep 06, 2011 at 10:52:41PM +0100, Mathew Newton wrote:
Well it's always good to know if ntpd falls out of sync.
As regards delivery of alerts by methods other than email, there are
these feature requests:
SMS alerting method
https://tools.bitfolk.com/redmine/issues/19
Twitter/Identica alerting method
https://tools.bitfolk.com/redmine/issues/16
They're ranked quite highly in votes so I will work on these soon.
Any comments on the tracker as to how you might like them to work
would be appreciated.
Cheers,
Andy
On Tue, September 6, 2011 10:33 pm, Andy Smith wrote:
Can I ask you to try both of:
- Installing kernel linux-image-2.6-686-bigmem
- Remove "clocksource=jiffies" from your kernel command line
You will need to "update-grub" after editing /boot/grub/menu.lst.
We have noticed timing oddities for the "jiffies" clocksource on the
new servers, which run a newer version of Xen compared to other
servers. The default clocksource is "xen".
I suspect your problems will go away once you do this.
- Which kernel you're running
2.6.26-2-xen-686
- What its command line is (look in grub/menu.lst
or /proc/cmdline)
/boot/vmlinuz-2.6.26-2-xen-686 root=UUID=35947b19-56f4-47c4-bc13
-3a7fe9c6b02a console=hvc0 ro clocksource=jiffies Would you like
us to monitor your ntpd for sync? This is a standard
Nagios check.
That would ordinarily be useful however as my e-mail sits on this server,
and the IMAP server falls over when the timing goes awry, it might not
help in this instance! If you think it might still pick up on more subtle
issues though then enabling it would be appreciated - thank you.