https://bugzilla.novell.com/show_bug.cgi?id=685276 https://bugzilla.novell.com/show_bug.cgi?id=685276#c8 --- Comment #8 from Roberto Scudeller <beto.rvs@gmail.com> 2011-04-11 16:25:29 UTC --- (In reply to comment #7)
(In reply to comment #6)
I run 53 VMs and configure a cronjob every 3 hours to "xm reboot" this VMs.
That doesn't answer the question: Is it important to have 40 (or 53, or any other particular number of) VMs, or is this reproducible also with just a single VM? Similarly - does rebooting after exactly 3 hours really matter? Is the rebooting part of the description relevant at all?
Doesn't matter the number of VMs, as long as I keep low load average, this problem don't appear. When I run a numerous VMs, with high load average (VM linux run CPU and IO benchmarks for example), this BUG appear.
iptables for a anti-spoof filter and basics rules (aloow >1024 tcp ports for example, with statefull).
Again, for me this doesn't answer the question: It's still unclear what, if any, relation iptables has to a crash in the blktap driver.
Finally, still waiting for the hypervisor log (#4 and #5 provide xend logs, which we unlikely will need here).
What other "hypervisor log" are needed? I send messages, xend-debug and xend.log. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.