Bug 1055064

Summary: targetcli restoreconfig doesn't restore all iSCSI portals
Product: Red Hat Enterprise Linux 7 Reporter: bfcutler
Component: kernelAssignee: Andy Grover <agrover>
Status: CLOSED CURRENTRELEASE QA Contact: Bruno Goncalves <bgoncalv>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 7.0CC: bdonahue, bfcutler, bgoncalv, svenkatr
Target Milestone: rc   
Target Release: ---   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: kernel-3.10.0-105.el7 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-06-13 12:27:13 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
write.sh reproduces the problem and write.log shows the error
none
updated test script none

Description bfcutler 2014-01-18 14:20:41 UTC
Created attachment 852002 [details]
write.sh reproduces the problem and write.log shows the error

Description of problem:

"targetcli restoreconfig" doesn't always restore all iSCSI portals.  This can happen on reboot but also can be triggered manually.


Version-Release number of selected component (if applicable):

Current RHEL7 Beta as of 1/18/14.

uname -r: 3.10.0-54.0.1.el7.x86_64

Version 2.1.fb31 of targetcli (per yum info targetcli)


How reproducible:

The attached tar file contains a script (write.sh) and results (write.log).

Simply execute "bash write.sh" and examine the output.  Upon failure, the output will contain one or more lines of the form:

Creating NetworkPortal object 192.168.1.2:3260 failed, skipped

Those lines should not be present and all portals should be correctly restored.

The example log provided has two such failures.


Additional info:

RHEL7 Beta is running within a VirtualBox (4.3.6) VM.  The host processor is a Xeon IvyBridge.  The host OS is Windows 7 Ultimate.

You can greatly increase the chance of failure by running with a single core.  Yes, I know that's not a supported configuration but perhaps it's helpful for debugging.

Comment 2 Andy Grover 2014-01-20 23:39:07 UTC
can reproduce

Comment 3 Andy Grover 2014-01-22 01:05:38 UTC
Created attachment 853546 [details]
updated test script

here's a cut-down version of the test script. Creating portals too fast leads to errors.

Comment 4 Andy Grover 2014-01-25 00:19:59 UTC
fix posted upstream

Comment 5 Andy Grover 2014-01-31 22:53:17 UTC
accepted upstream, waiting to show up in -stable

Comment 7 Jarod Wilson 2014-02-28 17:01:56 UTC
A kernel patch was submitted against this bug, so this bug needs to be against the kernel component. You'll have to reacquire flags now, unfortunately.

Comment 8 Jarod Wilson 2014-03-05 18:36:17 UTC
Patch(es) available on kernel-3.10.0-105.el7

Comment 11 Bruno Goncalves 2014-03-10 14:04:50 UTC
Verified. On kernel-3.10.0-105.el7.x86_64 it was not reproducible.

Easily reproducible on kernel-3.10.0-84.el7.x86_64

Configuration restored, 1 recoverable errors:
Creating NetworkPortal object 192.168.1.2:3260 failed: Could not create NetworkPortal in configFS., skipped


Note: I could reproduce this issue only on virtual machine.

Comment 12 Ludek Smid 2014-06-13 12:27:13 UTC
This request was resolved in Red Hat Enterprise Linux 7.0.

Contact your manager or support representative in case you have further questions about the request.