[Bug 425999] New: System Hang on a moderate IO load under XEN enabled kernel and memory over 10GB
https://bugzilla.novell.com/show_bug.cgi?id=425999 Summary: System Hang on a moderate IO load under XEN enabled kernel and memory over 10GB Product: openSUSE 10.3 Version: Final Platform: x86-64 OS/Version: openSUSE 10.3 Status: NEW Severity: Major Priority: P5 - None Component: Xen AssignedTo: cgriffin@novell.com ReportedBy: p.chiu@rl.ac.uk QAContact: qa@suse.de Found By: Customer We have recently bought 4 servers each with dual Intel E5440 quadcore cpus, 32GB memory, Areca ARC-1220 PCI-Express SATA RAID controller and 6 x 1TB WD WD1000FYPS disks. 2 of the 1TB SATA drives are set up in RAID-1, and 4 others in RAID-5. The RAID-1 storage is partitioned into / (100GB), /spare (100GB), swap (50GB), and /data (remaining 700GB in xfs). They are all installed with OpenSUSE 10.3 x86-64 and patched with the latest updates from download.opensuse.org. 3 of them are enabled with XEN. With xen enabled kernel running (2.6.22.18-0.2-xen x86_64), a simple dd command, eg. dd if=/dev/zero of=/data/dummy bs=1024k count=10000 to make a 10GB file will cause the system to hang a few minutes later. This problem happens on all the 3 servers with xen enabled kernel running, but NOT on the non-xen kernel running server. There is no error recorded in the system log. Furthermore, by trial and error, I found out that by reducing the physical memory size on the server to 4GB, the server seems to be quite happy to complete the dd and the subsequent bonnie++ tests targetted to it. I shall be grateful for any insight to explain this problem and the offer of a proper solution. -- 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.
https://bugzilla.novell.com/show_bug.cgi?id=425999
Jason Douglas
https://bugzilla.novell.com/show_bug.cgi?id=425999
User jbeulich@novell.com added comment
https://bugzilla.novell.com/show_bug.cgi?id=425999#c1
Jan Beulich
https://bugzilla.novell.com/show_bug.cgi?id=425999 User p.chiu@rl.ac.uk added comment https://bugzilla.novell.com/show_bug.cgi?id=425999#c2 Peter Chiu
https://bugzilla.novell.com/show_bug.cgi?id=425999
User jbeulich@novell.com added comment
https://bugzilla.novell.com/show_bug.cgi?id=425999#c3
Jan Beulich
https://bugzilla.novell.com/show_bug.cgi?id=425999 User p.chiu@rl.ac.uk added comment https://bugzilla.novell.com/show_bug.cgi?id=425999#c4 Peter Chiu
https://bugzilla.novell.com/show_bug.cgi?id=425999
User jbeulich@novell.com added comment
https://bugzilla.novell.com/show_bug.cgi?id=425999#c5
Jan Beulich
https://bugzilla.novell.com/show_bug.cgi?id=425999 User p.chiu@rl.ac.uk added comment https://bugzilla.novell.com/show_bug.cgi?id=425999#c6 --- Comment #6 from Peter Chiu
https://bugzilla.novell.com/show_bug.cgi?id=425999
User jdouglas@novell.com added comment
https://bugzilla.novell.com/show_bug.cgi?id=425999#c7
--- Comment #7 from Jason Douglas
Can you please be more explicit as to how to collect the kernel and hypervisor messages you need.
You mentioned about the serial cable, and the sync_console command, can you elaborate in details how to do that.
Instructions for collecting kernel and hypervisor messages can be found here: http://en.opensuse.org/How_to_Capture_Xen_Hypervisor_and_Kernel_Messages_usi... There are no instructions specifically geared towards openSUSE 10.3, but one of the two examples at the bottom of the page should probably work. -- 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.
participants (1)
-
bugzilla_noreply@novell.com