Bug 1657672
Summary: | [Marvell 7.6 Bug] Unnecessary Login attempts are done even after successful login, causing delay in system boot | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | Red Hat Enterprise Linux 7 | Reporter: | Nilesh Javali <nilesh.javali> | ||||||
Component: | iscsi-initiator-utils | Assignee: | Chris Leech <cleech> | ||||||
Status: | CLOSED WONTFIX | QA Contact: | Martin Hoyer <mhoyer> | ||||||
Severity: | high | Docs Contact: | |||||||
Priority: | unspecified | ||||||||
Version: | 7.6 | CC: | cleech, fsuba, girish.basrur, manish.rangankar, nilesh.javali, njavali | ||||||
Target Milestone: | rc | ||||||||
Target Release: | --- | ||||||||
Hardware: | x86_64 | ||||||||
OS: | Linux | ||||||||
Whiteboard: | |||||||||
Fixed In Version: | Doc Type: | If docs needed, set a value | |||||||
Doc Text: | Story Points: | --- | |||||||
Clone Of: | |||||||||
: | 1658483 (view as bug list) | Environment: | |||||||
Last Closed: | 2021-03-15 07:32:18 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: | |||||||||
Bug Depends On: | |||||||||
Bug Blocks: | 1658483 | ||||||||
Attachments: |
|
Below are the multipath.conf settings, cat /etc/multipath.conf defaults { find_multipaths yes user_friendly_names yes polling_interval 5 fast_io_fail_tmo 5 dev_loss_tmo 10 checker_timeout 15 failback immediate } Chris, We are seeing this issue in RHEL 8.0 as well. However looking at the iscsiroot.sh script (/usr/lib/dracut/modules.d/95iscsi/) the method to initiate login to iSCSI target has changed from "iscsistart -b" to "iscsiadm -m fw -l". And any of the rd.iscsi.param is not honored. Can you please check. Chris, In cases of link down on one port, does the script check for the link state before triggering login attempts and can that be controlled. Created attachment 1687657 [details]
RHEL 8.2 logs
Hi Chris,
We are facing similar issue on RHEL 8.2 iSCSI boot from SAN multipath with one port down setup. It looks like Dracut login retries is adding significant delay in boot process when one port is down.
When the one is down and system rebooted dracut retries for the iSCSI login 12 times. Which takes approximately 22 min starting from Apr 23 11:17:48 to 11:39:04. Which is a significant portion of the boot process. Refer below log snippet for duration.
==================
338152:Apr 23 11:17:41 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338232:Apr 23 11:19:34 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338273:Apr 23 11:21:20 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338314:Apr 23 11:23:10 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338355:Apr 23 11:24:56 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338396:Apr 23 11:26:43 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338437:Apr 23 11:28:32 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338478:Apr 23 11:30:16 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338519:Apr 23 11:32:03 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338560:Apr 23 11:33:49 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338601:Apr 23 11:35:33 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338642:Apr 23 11:37:19 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
338683:Apr 23 11:39:04 RH82-rc-router dracut-initqueue[1317]: iscsiadm: Could not login to [iface: qedi.00:0e:1e:d2:dd:80, target: iqn.2001-05.com.equallogic:0-8a0906-5fe6d7608-ae60a2f9a5b5e98a-rh82-rc, portal: 192.168.91.15,3260].
=================
Is there a way to reduce Dracut retires using cmdline parameter ?
Chris, any updates on this? After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened. |
Created attachment 1512971 [details] RH7.6 OS boot logs Description of problem: In iSCSI offload Boot from SAN environment with multipath configuration, during OS boot the iscsistart unnecessarily attempts to login to already logged in sessions thus causing delay in system boot. Version-Release number of selected component (if applicable): RHEL 7.6 GA How reproducible: Always 100% reproducible Steps to Reproduce: 1. Install RHEL 7.6 GA OS on iSCSI LUN using QLogic Fastlinq CNA adapter via iSCSI offload path in mutipath configuration. 2. The OS installs successfully 3. On the first OS boot it is observed that iscsistart continuously attempts login to iSCSI sessions already being established. [ 81.531880] dracut-initqueue[2342]: iscsistart: version 6.2.0.874-10 [ 82.535366] dracut-initqueue[2342]: iscsistart: initiator reported error (15 - session exists) [ 82.540439] dracut-initqueue[2342]: iscsistart: initiator reported error (15 - session exists) [ 82.545569] dracut-initqueue[2342]: Warning: 'iscsistart -b ' failed with return code 0 4. This adds delay in booting of OS. Actual results: The OS boot takes more than 5 mins to boot. Expected results: The OS should boot within less than 5 mins. Additional info: Adding OS boot logs showing multiple login attempts.