Description of problem: Last couple of months we have been working on to add multiple patches to scsi rdac device handler. All these patches are reviewed and approved. It will be available in 2.6.32 kernel. I am opening this bugzilla to include all these patches in RHEL 5.5 whenever the window open up. These patches add mailnly 3 features. a) First three patches include new device in rdac handler b) One patch fixes a potential bug. b) Next three patches include one the fly debug functionality which is very important for us. Here are the details of the patches 1. Add support for Sun StorageTek ST2500, ST2510 and ST2530 http://marc.info/?l=linux-scsi&m=125132741303628&w=2 2. Add support for two more SUN devices http://marc.info/?l=linux-scsi&m=125053849203620&w=2 3. Add support for next generation of Dell PowerVault storage array. http://marc.info/?l=dm-devel&m=125055416131006&w=2 4. Fix for returning correct mode select cmd return info http://marc.info/?l=dm-devel&m=125148188605897&w=3 5. Following three patches include debugging functinalities to rdac handler. http://marc.info/?l=linux-scsi&m=125242415500674&w=3 http://marc.info/?l=linux-scsi&m=125242417300729&w=3 http://marc.info/?l=linux-scsi&m=125242421700850&w=3 I am porting all these patches on 2.6.18-164.el5. I will attach the Patches ASAP.
Created attachment 361724 [details] Add support for Sun StorageTek ST2500, ST2510 and ST2530
Created attachment 361725 [details] Add support for two more SUN devices
Created attachment 361726 [details] Add support for next generation of Dell PowerVault storage array
Created attachment 361727 [details] Fix for returning correct mode select cmd return info
Created attachment 361728 [details] [PATCH 1/3][RESUBMIT] scsi_dh_rdac: move the init code from rdac_activate to rdac_attach
Created attachment 361729 [details] [PATCH 2/3][RESUBMIT] scsi_dh_rdac: changes to collect the rdac debug information during the initialization
Created attachment 361730 [details] [PATCH 3/3][RESUBMIT] scsi_dh_rdac: changes for rdac debug logging
Created attachment 361731 [details] [PATCH 3/3][RESUBMIT] scsi_dh_rdac: changes for rdac debug logging
*** Bug 520860 has been marked as a duplicate of this bug. ***
I tried to apply the first attachement patch and it applied with fuzz to today's rhel5.5. Babu - Please create an ordered patch set that applies cleanly to rhel5.4. Andrius, can you provide a link to rhel5.4 for Babu?
Babu, Can you also comment on any testing that was done to validate patches in attachments in 4, and set 5,6, and 8. Thanks, Rob
Yes, I have tested these patches on 2.6.31 kernel with LSI rdac storage(base os RHEL 5.3). Patches work fine. As you know these patches add more rdac specific debug messages during the failures. Following are test details.. Configuration : LSI rdac storage with 4 luns. I had 4 paths to each lun(2 active and 2 passive). Ran overnight failover/failback tests. Tests ran fine. Enabled the logging using module parameter rdac_logging. Noticed failover/failback messages in /var/log/messages file. Without these patches, we will not see rdac specific messages during the failures and becomes very difficult to debug.
Babu, Unfortunately the testing of these patches w/ the 2.6.31 kernel is not sufficient for inclusion into a rhel release, especially one that is branched from 2.6.18. Please test again against rhel5.4 and describe the testing that was done to validate the attachments in 4, and set 5,6, and 8. Thanks, Rob
Rob, I regenerated all the patches on top of 2.6.18-164 and tested it. Patches work fine. Here are details. Configuration : LSI rdac storage with 2 luns with 2 paths to each lun(1 active and 1 passive). Verified lun discovery. Ran IO with failover/failback tests. Tests ran fine. Enabled/disabled logging using module parameter rdac_logging. Noticed failover/failback messages in /var/log/messages file. Without these patches, we will not see rdac specific messages during the failures and becomes very difficult to debug. Tests are still running. Will report back if I see any issues. -Thanks, Babu
Created attachment 362724 [details] Add support for Sun StorageTek ST2500, ST2510 and ST2530
Created attachment 362725 [details] Add support for two more SUN devices
Created attachment 362726 [details] Add support for next generation of Dell PowerVault storage array
Created attachment 362727 [details] Fix for returning correct mode select cmd return info
Created attachment 362728 [details] [PATCH 1/3][RESUBMIT] scsi_dh_rdac: move the init code from rdac_activate to rdac_attach
Created attachment 362729 [details] [PATCH 2/3][RESUBMIT] scsi_dh_rdac: changes to collect the rdac debug information during the initialization
Created attachment 362730 [details] [PATCH 3/3][RESUBMIT] scsi_dh_rdac: changes for rdac debug logging
Thanks Babu, This patchset applied cleanly to the current rhel5.5 tree and built without issue. Rob
Rob, Thanks for the confirmation. -Babu
*** Bug 526241 has been marked as a duplicate of this bug. ***
in kernel-2.6.18-169.el5 You can download this test kernel from http://people.redhat.com/dzickus/el5 Please do NOT transition this bugzilla state to VERIFIED until our QE team has sent specific instructions indicating when to do so. However feel free to provide a comment indicating that this fix has been verified.
I have verified the patches and also tested the kernel 2.6.18-169.el5. Everything is working as expected. Configuration: LSI storage with scsi rdac handler(with 2 luns with 2 paths(one active and one passive) Kernel version 2.6.18-169.el5. Tested - failover/failback successfully. Also verified logs to see the messages are logged as expected. -thanks, Babu
Technical note added. If any revisions are required, please edit the "Technical Notes" field accordingly. All revisions will be proofread by the Engineering Content Services team. New Contents: Added rdac support for the following devices: Sun StorageTek ST2500 Sun StorageTek ST2510 Sun StorageTek ST2530 Sun CSM100_R_FC Dell PowerVault MD32xx Dell PowerVault MD32xxi A bug was fixed where incorrect sense data needed to be flagged as present to upper layers when not handled by the rdac device handler.
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. http://rhn.redhat.com/errata/RHSA-2010-0178.html