Bugs item #2723366, was opened at 2009-03-31 07:00 Message generated for change (Comment added) made by jiajun You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=893831&aid=2723366&group_id=180599 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: qemu Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Jiajun Xu (jiajun) Assigned to: Nobody/Anonymous (nobody) Summary: ltp diotest running time is 2.54 times than before Initial Comment: Running LTP diotest in guest costs about 2m22s in latest commit which it needs 0m56s in older commit. New kvm kernel commit:a317a1e496b22d1520218ecf16a02498b99645e2 kvm user-space commit: df0e52a8d988d55dd42f8d46faffa9faa41892c9 ltp diotest running time: 5 rounds of diotest average time: 2m22s Old kvm kernel commit:e74bb3fa8e55284dc6fdd68aa9da833ce07a4295 kvm user-space commit: 4c1083fd610fba4eedb45553c0a579b7b6593f1a ltp diotest running time: 5 rounds of diotest average time: 0m56s Reproduce steps: (1)qemu-system-x86_64 -m 512 -smp 4 -net nic,macaddr=00:16:3e:12:0d:3c,model=rtl8139 -net tap,script=/etc/kvm/qemu-ifup -hda /share/xvs/var/ia32e_rhel4u1.img (2)Get LTP package from sourceforge and install it in guest (3)Run diotest for 5 times. while [ $i -lt 5 ] do time /ltp-full-20070930/runltp -l /tmp/ltp_dio.log -f dio -p -q >> /root/ltp.log ((i++)) done Notes: The data is running with ltp-full-20070930, and we also tried latest ltp package ltp-full-20090228, which can get similar data. ---------------------------------------------------------------------- >Comment By: Jiajun Xu (jiajun) Date: 2009-10-18 22:55 Message: I checked with kvm: f3ed8d80..., qemu-kvm: 3c62010..., is still takes more than 2 minutes to finish diotest with "cache=writethrough". ---------------------------------------------------------------------- Comment By: Marcelo Tosatti (mtosatti) Date: 2009-10-17 18:10 Message: Jiajun, This was probably caused by the default caching mode change to writeback. Can you verify if "cache=writethrough" option takes performance back to 0m56s range as before? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=893831&aid=2723366&group_id=180599 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
