Hello,
On Sat, Jun 11, 2011 at 10:18:10PM +0000, Andy Smith wrote:
It appears we've hit some sort of bug here and
customers aren't able
to start new virtual machines on this host. Therefore I'm going to
have to reboot the host now as a matter of urgency, i.e. in the next
few minutes.
Reboot was started at approx. 2221Z and all VPSes had booted again
by approx. 2254Z.
There was a period of high IO load between ~2230Z and ~2330Z as many
customers on this host had not rebooted for hundreds of days and
filesystem checks all kicked in at once. This has now returned to
normal.
http://tools.bitfolk.com/cacti/graph_1113.html
I do not expect to have to take this host out of service in the near
future, and the forced reboot and subsequent upgrade to 64bit
kernel/hypervisor has allowed the host to address additional memory
beyond the 16G it was limited to before.[*] This has removed the need
for a planned reboot that would have been scheduled for sometime in
the next two weeks.
Thank you for your patience.
Cheers,
Andy
[*] I am loving the footnotes tonight. 32bit Xen can only address up
to 16G RAM in PAE mode, which was a bit of a surprise to me at
the time as Linux PAE implementation can address up to 64G.
--
http://bitfolk.com/ -- No-nonsense VPS hosting