Red Hat Bugzilla – Bug 623595
move iscsi/iser to passthrough mode, fix functioning and failover time under DM multipath
Last modified: 2011-02-25 05:54:52 EST
Created attachment 438383 [details]
patch which has the related fixes from upstream
Description of problem:
Remove some inefficiencies in the iser data path through
simplification and reducing the amount of code, using less atomic
operations, avoiding TX interrupts, moving to iscsi passthrough mode, etc.
As a result I was able to duplicate the amount of IOPS, upto about 300K
and to reduce the latency measured through iostat AWAIT report.
The upstream submission was done through a sequence of patches
and not as one big re-write, to ease with maintainance and tuning.
Fix bugs in the disconnection logic of iser which made the failover time with dm-multipathing to be too long, at least 120 seconds, wheres now it can get to be 30 seconds and below that. This part is the same as RHEL6 bugzilla 589174
Attached is the patch which addresses all that.
Forgot to mention that the patch was generated against / tested with kernel 2.6.18-211.el5
This seems fine to me.
It is upstream, Voltaire is doing testing, and they made the patch for us. Adding devel ack.
This request was evaluated by Red Hat Product Management for inclusion in a Red
Hat Enterprise Linux maintenance release. Product Management has requested
further review of this request by Red Hat Engineering, for potential
inclusion in a Red Hat Enterprise Linux Update release for currently deployed
products. This request is not yet committed for inclusion in an Update
You can download this test kernel from http://people.redhat.com/jwilson/el5
Detailed testing feedback is always welcomed.
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.