Bug 391951
| Summary: | [NetApp 5.2 bug] root on multipath (iSCSI) not supported in Anaconda | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 5 | Reporter: | Ritesh Raj Sarraf <rsarraf> | ||||||||
| Component: | anaconda | Assignee: | David Cantrell <dcantrell> | ||||||||
| Status: | CLOSED ERRATA | QA Contact: | Alexander Todorov <atodorov> | ||||||||
| Severity: | high | Docs Contact: | |||||||||
| Priority: | high | ||||||||||
| Version: | 5.1 | CC: | andriusb, coughlan, k.georgiou, mchristi, nandkumar.mane, poelstra, xdl-redhat-bugzilla | ||||||||
| Target Milestone: | --- | Keywords: | OtherQA | ||||||||
| Target Release: | --- | ||||||||||
| Hardware: | All | ||||||||||
| OS: | Linux | ||||||||||
| Whiteboard: | |||||||||||
| Fixed In Version: | RHBA-2008-0397 | Doc Type: | Bug Fix | ||||||||
| Doc Text: | Story Points: | --- | |||||||||
| Clone Of: | Environment: | ||||||||||
| Last Closed: | 2008-05-21 15:32:20 UTC | Type: | --- | ||||||||
| Regression: | --- | Mount Type: | --- | ||||||||
| Documentation: | --- | CRM: | |||||||||
| Verified Versions: | Category: | --- | |||||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||||
| Embargoed: | |||||||||||
| Bug Depends On: | |||||||||||
| Bug Blocks: | 217208, 428723 | ||||||||||
| Attachments: |
|
||||||||||
|
Description
Ritesh Raj Sarraf
2007-11-20 11:44:23 UTC
Created attachment 264841 [details]
anaconda iscsi multipath patch
Attached is a patch which fixes the problem described above. We are able to see
iSCSI devices as multipathed devices and do a root installation on an iSCSI
multipathed device.
But it introduces a new problem. On reboot, the kernel panics because it can't
find the multipathed device. From the logs on the screen, we found that the
network and iSCSI services never started when executing in initrd. Looks like
the post-installation of anaconda is screwing up something.
But a normal iSCSI device based installation is proper and a working initrd
image is generated.
We need help in understanding how anaconda generates the initrd image and what
all dependencies are involved in it.
Any help/pointers will be appreciated.
Adding andrius and coughlan to cc list. Can we have some feedback on this patch? We'd like to see iSCSI root on multipath support in RHEL5.2. With this patch applied, anaconda will be able to list iSCSI devices as multipathed devices and proceed with installation of '/' on an iSCSI LUN. (In reply to comment #1) > Created an attachment (id=264841) [edit] > anaconda iscsi multipath patch > > Attached is a patch which fixes the problem described above. We are able to see > iSCSI devices as multipathed devices and do a root installation on an iSCSI > multipathed device. > > But it introduces a new problem. On reboot, the kernel panics because it can't > find the multipathed device. From the logs on the screen, we found that the > network and iSCSI services never started when executing in initrd. Looks like > the post-installation of anaconda is screwing up something. > > But a normal iSCSI device based installation is proper and a working initrd > image is generated. > > We need help in understanding how anaconda generates the initrd image and what > all dependencies are involved in it. > Any help/pointers will be appreciated. The patch provided seems to only address half of the problem, which you point out. The root multipath support in RHEL 5.1 is implemented in anaconda for installation support and also in mkinitrd for initrd.img image generation. For iSCSI root multipath support, you'll need to also check the mkinitrd package and patch as necessary to complete the support. In the mkinitrd SRPM for RHEL 5.1, you'll find this patch: mkinitrd-5.1.19.6-multipath.patch That's all of the root multipath work. We'll probably need to modify the device gathering functions so we write out the correct nash script to bring the system up. Created attachment 291510 [details] iscsi mkinitrd patch Attached is a patch which solves the initrd image corruption issue. Now we are able to see multipathed device as a root. Previous anaconda attachment( attachment id=264841 (comment #1)) had solved installation issue over multipathed device. But it had introduced new problem. On reboot, the kernel panics because it can't find the multipathed device. Here anaconda post installation was screwing up initrd image, only in case of multipathed device. mkinitrd patch makes sure that multipathed devices get handled properly. These patches look good. I'm cloning this bug for mkinitrd since we need the second patch to go in to that component. Thanks David. Can we have update on this bug? (In reply to comment #9) > Thanks David. > Can we have update on this bug? The anaconda patch in comment #1 is included in the 5.2 beta release. The mkinitrd patch in comment #6 is also in, but that is tracked by BZ 428723, since it is a different component. NetApp, can you verify that the fix is included in anaconda and it works? Thanks! Thank you. Will do once the betas are released. Is there any other channel to get early access to the packages committed for RHEL5.2? I've been tracking the Anaconda git archive at Fedora but I don't see the changes there. NetApp, as RHEL 5.2 beta is out now can you please verify if this issue is fixed. Thanks Created attachment 297753 [details]
mkinitrd patch for root on multipath(iscsi) feature
anaconda patch has been included in 5.2 Beta.(works fine).
Though mkintrd patch has been included in mkintrd src, it doesn't apply during
package building. The switch (-p0) was supposed to be used to apply the patch
cleanly.
Attached is the same patch with the -p1 option.
Yes, this is a problem. I have moved the problem report to BZ 428723, though, because that is the BZ for the mkinitrd patch. This one is for Anaconda. Greetings Red Hat Partner, A fix for this issue should be included in the latest packages contained in RHEL5.2-Snapshot1--available now on partners.redhat.com. Please test and confirm that your issue is fixed. After you (Red Hat Partner) have verified that this issue has been addressed, please perform the following: 1) Change the *status* of this bug to VERIFIED. 2) Add *keyword* of PartnerVerified (leaving the existing keywords unmodified) If this issue is not fixed, please add a comment describing the most recent symptoms of the problem you are having and change the status of the bug to ASSIGNED. If you are receiving this message in Issue Tracker, please reply with a message to Issue Tracker about your results and I will update bugzilla for you. If you need assistance accessing ftp://partners.redhat.com, please contact your Partner Manager. Thank you mkinitrd patch has not been included in RHEL 5.2 Snapshot1. Patch from Comment #17 is the valid patch for the mkinitrd problem. Setting back to ON_QA becuase the issue found in Comment #21 has moved to bug 428723. Greetings Red Hat Partner, A fix for this issue should be included in the latest packages contained in RHEL5.2-Snapshot3--available now on partners.redhat.com. Please test and confirm that your issue is fixed. After you (Red Hat Partner) have verified that this issue has been addressed, please perform the following: 1) Change the *status* of this bug to VERIFIED. 2) Add *keyword* of PartnerVerified (leaving the existing keywords unmodified) If this issue is not fixed, please add a comment describing the most recent symptoms of the problem you are having and change the status of the bug to ASSIGNED. If you are receiving this message in Issue Tracker, please reply with a message to Issue Tracker about your results and I will update bugzilla for you. If you need assistance accessing ftp://partners.redhat.com, please contact your Partner Manager. Thank you verified anaconda patch is present in anaconda-11.1.2.111-1 ... also verified the patch for bug 428723 is present in mkinitrd-5.1.19.6-28 ... do we have any results from our partner that the multipath behavior is verified? The fix is included in the latest snapshot of RHEL5.2 and has been verified. With this patch in place, iSCSI devices are listed as multipathed devices and installation on those devices occurs properly. Greetings Red Hat Partner, A fix for this issue should be included in the latest packages contained in RHEL5.2-Snapshot4--available now on partners.redhat.com. Please test and confirm that your issue is fixed. After you (Red Hat Partner) have verified that this issue has been addressed, please perform the following: 1) Change the *status* of this bug to VERIFIED. 2) Add *keyword* of PartnerVerified (leaving the existing keywords unmodified) If this issue is not fixed, please add a comment describing the most recent symptoms of the problem you are having and change the status of the bug to ASSIGNED. If you are receiving this message in Issue Tracker, please reply with a message to Issue Tracker about your results and I will update bugzilla for you. If you need assistance accessing ftp://partners.redhat.com, please contact your Partner Manager. Thank you I have a requirement to use iscsi root multipath and during some testing with the anaconda patch (under 5.1) I am not sure that it is complete. * anaconda doesn't mark the root device as _netdev it should be _rnetdev after5.2 though. * I was unable to use lvm with it under ks part pv.00 --size=14000 --grow --ondevice=mapper/mpath0 fails but this could be because I have no idea on how the multipath kickstart command works and I was unable to find any docs for it. Yeah. I agree with your first point. Currently anaconda is not marking root multipath device as _netdev. Will fix this in 5.3. 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 the 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/RHBA-2008-0397.html |