Red Hat Bugzilla – Bug 636291
[LSI 6.1 bug] RHEL 6.0 iSCSI offload (cxgb3i) sessions do not log back in after several controller reset cycles [LSI CR184419]
Last modified: 2012-04-20 12:30:28 EDT
Created attachment 448802 [details] host messages log Description of problem: RHEL 6.0 iSCSI offload (cxgb3i) sessions do not log back in after several controller reset cycles. Problem recreated with read error at 23 minutes in one test, and read error at 41 minutes in another test. Error logs show both read and write errors with multiple volumes on both arrays. This is similar issue to bug 567444 which we found in RHEL5.5. Version-Release number of selected component (if applicable): Kernel Release: 2.6.32-66.el6.ppc64 RHEL Release: Red Hat Enterprise Linux Server release 6.0 Beta (Santiago) Version: Linux version 2.6.32-66.el6.ppc64 (mockbuild@js20-bc2-10.build.redhat.com) (gcc version 4.4.4 20100726 (Red Hat 4.4.4-13) (GCC) ) #1 SMP Wed Aug 18 01:12:05 EDT 2010 Platform: ppc64 MPP Version: Linux MPP Driver Version: 99.03.0C05.0427 Arrays Connected: --------------------------------------------------------------- Info of Array Module's seen by this Host. --------------------------------------------------------------- ID WWN Type Name --------------------------------------------------------------- --------------------------------------------------------------- Mapped LUNS: 0 device-mapper-event-libs-1.02.53-8.el6.ppc64 device-mapper-event-1.02.53-8.el6.ppc64 device-mapper-1.02.53-8.el6.ppc64 device-mapper-multipath-libs-0.4.9-28.el6.ppc64 device-mapper-multipath-0.4.9-28.el6.ppc64 device-mapper-libs-1.02.53-8.el6.ppc64 driver: cxgb3 version: 1.1.3-ko firmware-version: T 7.11.0 TP 1.1.0 bus-info: 0004:01:00.0 driver: cxgb3 version: 1.1.3-ko firmware-version: T 7.11.0 TP 1.1.0 bus-info: 0005:01:00.0 cxgb3: filename: /lib/modules/2.6.32-66.el6.ppc64/kernel/drivers/net/cxgb3/cxgb3.ko firmware: cxgb3/ael2020_twx_edc.bin firmware: cxgb3/ael2005_twx_edc.bin firmware: cxgb3/ael2005_opt_edc.bin firmware: cxgb3/t3c_psram-1.1.0.bin firmware: cxgb3/t3b_psram-1.1.0.bin firmware: cxgb3/t3fw-7.4.0.bin version: 1.1.3-ko license: Dual BSD/GPL author: Chelsio Communications description: Chelsio T3 Network Driver srcversion: D21F8F820FC4726C89DA2AC alias: pci:v00001425d00000037sv*sd*bc*sc*i* alias: pci:v00001425d00000036sv*sd*bc*sc*i* alias: pci:v00001425d00000035sv*sd*bc*sc*i* alias: pci:v00001425d00000032sv*sd*bc*sc*i* alias: pci:v00001425d00000031sv*sd*bc*sc*i* alias: pci:v00001425d00000030sv*sd*bc*sc*i* alias: pci:v00001425d00000026sv*sd*bc*sc*i* alias: pci:v00001425d00000025sv*sd*bc*sc*i* alias: pci:v00001425d00000024sv*sd*bc*sc*i* alias: pci:v00001425d00000023sv*sd*bc*sc*i* alias: pci:v00001425d00000022sv*sd*bc*sc*i* alias: pci:v00001425d00000021sv*sd*bc*sc*i* alias: pci:v00001425d00000020sv*sd*bc*sc*i* depends: mdio vermagic: 2.6.32-66.el6.ppc64 SMP mod_unload modversions parm: dflt_msg_enable:Chelsio T3 default message enable bitmap (int) parm: msi:whether to use MSI or MSI-X (int) parm: ofld_disable:whether to enable offload at init time or not (int) cxgb3i: filename: /lib/modules/2.6.32-66.el6.ppc64/kernel/drivers/scsi/cxgb3i/cxgb3i.ko version: 1.0.2 license: GPL description: Chelsio S3xx iSCSI Driver author: Karen Xie <kxie@chelsio.com> srcversion: 8B5AFD106B375820279FEC9 depends: libiscsi_tcp,scsi_transport_iscsi,libiscsi,cxgb3 vermagic: 2.6.32-66.el6.ppc64 SMP mod_unload modversions parm: cxgb3_rcv_win:TCP receive window in bytes (default=256KB) (int) parm: cxgb3_snd_win:TCP send window in bytes (default=128KB) (int) parm: cxgb3_rx_credit_thres:int parm: rx_credit_thres:RX credits return threshold in bytes (default=10KB) parm: cxgb3_max_connect:Max. # of connections (default=8092) (uint) parm: cxgb3_sport_base:starting port number (default=20000) (uint) How reproducible: Often. Steps to Reproduce: 1. Created 64 volumes from two storages. 2. Mapped 64 volumes from each array to the RHEL6 host. 3. Start the 4 cxgb3i session 2 for each array. 4. Start I/Os to all volumes from the host. 5. Reset one of the controllers of each array. Actual results: The host with iSCSI offload(cxgb3i) failed to log back to the array after it completed reset. Expected results: The session log back to the controller. Additional info: This is similar issue to bug 567444 which we found in RHEL5.5.
(In reply to comment #0) > Created attachment 448802 [details] > host messages log > > Description of problem: > > RHEL 6.0 iSCSI offload (cxgb3i) sessions do not log back in after several > controller reset cycles. Problem recreated with read error at 23 minutes in one > test, and read error at 41 minutes in another test. Error logs show both read > and write errors with multiple volumes on both arrays. > > This is similar issue to bug 567444 which we found in RHEL5.5. > That bz is a generic update driver to version xyz. Do you mean that the patches merged for bz 567444 fixed your issue in RHEL 5.5, or did you mean that the issue is still present with in RHEL 5.5 even with those patches?
We don't have any free RHEL5.5 setup at the moment, but will see if we can get something up by next week to verify the suggested patches from 567444.
LSI tested the kernel-2.6.18-225.el5 for RHEL5.5 from bug 567444 and passed all the failed tests before. When can we get the same (GA) patches for RHEL6.
Created attachment 454227 [details] Prevent garbage values from being used. Hi LSI, Could you try this patch? It is the only patch we are missing from the RHEL 5 code you tested. The patch should apply to the current RHEL 6 kernel (there might be some offsets, but that is fine).
Hi Mike; can we get a test kernel with that patch in it as we did for RHEL 5.5?
(In reply to comment #6) > Hi Mike; > can we get a test kernel with that patch in it as we did for RHEL 5.5? I will have to make one myself. Tell me what arch you are using so I only have to build the specific kernel you are using.
We are currently running with kernel 2.6.32-66.el6.ppc64.
I put a ppc64 kernel with the fix here: http://people.redhat.com/mchristi/iscsi/rhel6.0/kernel/test/kernel-2.6.32-79.el6.cxgb3ilogin.ppc64.rpm
This request was evaluated by Red Hat Product Management for inclusion in the current release of Red Hat Enterprise Linux. Because the affected component is not scheduled to be updated in the current release, Red Hat is unfortunately unable to address this request at this time. Red Hat invites you to ask your support representative to propose this request, if appropriate and relevant, in the next release of Red Hat Enterprise Linux. If you would like it considered as an exception in the current release, please ask your support representative.
(In reply to comment #10) > This request was evaluated by Red Hat Product Management for > inclusion in the current release of Red Hat Enterprise Linux. > Because the affected component is not scheduled to be updated > in the current release, Red Hat is unfortunately unable to > address this request at this time. Red Hat invites you to > ask your support representative to propose this request, if > appropriate and relevant, in the next release of Red Hat > Enterprise Linux. If you would like it considered as an > exception in the current release, please ask your support > representative. Please disregard this, as it was in error.
This request was erroneously denied for the current release of Red Hat Enterprise Linux. The error has been fixed and this request has been re-proposed for the current release.
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 release.
Patch(es) available on kernel-2.6.32-112.el6
~~ Partners and Customers ~~ This bug was included in RHEL 6.1 Beta. Please confirm the status of this request as soon as possible. If you're having problems accessing 6.1 bits, are delayed in your test execution or find in testing that the request was not addressed adequately, please let us know. Thanks!
*** Bug 647016 has been marked as a duplicate of this bug. ***
Created attachment 490376 [details] cxgb4i upstream commit ids.
Created attachment 490377 [details] backported cxgb4i to RHEL 6 kernel 2.6.32-71.7.1.
Created attachment 490378 [details] backported cxgb4i to RHEL 6 kernel 2.6.32-71.7.1.
@IBM, @Chelsio, @LSI, 6.1.0 Test results?
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-2011-0542.html
WRT to Hongs issue from LSI, iSCSI offload session are working fine in 2.6.32-257.el6 kernel. I have listed the modinfo for cxgb3i driver and uname for the system information root@ictm-ayeka Desktop]# modinfo cxgb3i filename: /lib/modules/2.6.32-257.el6.ppc64/kernel/drivers/scsi/cxgbi/cxgb3i/cxgb3i.ko license: GPL version: 2.0.0 description: Chelsio T3 iSCSI Driver author: Chelsio Communications, Inc. srcversion: 5CCD001CC4C03BDEE94C11C depends: libiscsi,libcxgbi,libiscsi_tcp,cxgb3 vermagic: 2.6.32-257.el6.ppc64 SMP mod_unload modversions parm: dbg_level:debug flag (default=0) (uint) parm: cxgb3i_rcv_win:TCP receive window in bytes (default=256KB) (int) parm: cxgb3i_snd_win:TCP send window in bytes (default=128KB) (int) parm: cxgb3i_rx_credit_thres:int parm: rx_credit_thres:RX credits return threshold in bytes (default=10KB) parm: cxgb3i_max_connect:Max. # of connections (default=8092) (uint) parm: cxgb3i_sport_base:starting port number (default=20000) (uint) [root@ictm-ayeka Desktop]# uname -a Linux ictm-ayeka 2.6.32-257.el6.ppc64 #1 SMP Mon Mar 26 10:17:40 EDT 2012 ppc64 ppc64 ppc64 GNU/Linux The session are logging back into the array after several controller reset. Thanks