iSCSI timeouts Q&A
Applies to
- Data ONTAP 8 7-Mode
- Data ONTAP 7 and earlier
Answer
Error message: [iscsi.notice:notice]: ISCSI: iswta, LUN Reset (from initiator iqn.1991-05.com.microsoft:exch), aborting all SCSI commands on lun 0
Error message:
[iscsi.notice:notice]: ISCSI: iswta, New session from initiator iqn.1991-05.com.microsoft:exch at IP addr 10.0.0.1
iScsiPrt Information None 34 N/A SOUTHPOINT
iScsiPrt Error None 39 N/A SOUTHPOINT
Initiator sent a task management command to reset the target. The target name is given in the dump data.
iScsiPrt Error None 9 N/A SOUTHPOINT
Target did not respond in time for a SCSI request. The CDB is given in the dump data.
A connection to the target was lost, but Initiator successfully reconnected to the target. Dump data contains the target name.
iSCSI Microsoft software initiator
- I have migrated from VLD to iSCSI. After migration, I see the following error messages in the Windows host Event Logs and the filer. I am using the Microsoft software initiator. However my application is running fine. Should I be concerned?
iScsiPrt Information None 34 N/A SOUTHPOINT
A connection to the target was lost, but Initiator successfully reconnected to the target. Dump data contains the target name.
iScsiPrt Error None 39 N/A SOUTHPOINT
Initiator sent a task management command to reset the target. The target name is given in the dump data.
iScsiPrt Error None 9 N/A SOUTHPOINT
Target did not respond in time for a SCSI request. The CDB is given in the dump data.
And you may see repeated messages like the followings in filer's syslog because of host-side SCSI timeout:
Error message:[iscsi.notice:notice]: ISCSI: iswta, LUN Reset (from initiator iqn.1991-05.com.microsoft:exch), aborting all SCSI commands on lun 0
Error message:[iscsi.notice:notice]: ISCSI: iswta, New session from initiator iqn.1991-05.com.microsoft:exch at IP addr 10.0.0.1
These error messages are logged when the Client side SCSI Storage stack tries to execute an I/O, and the I/O did not complete with a specified period of time (the default timeout value is 25 secs for Microsoft software initiator, See KB: Host-side SCSI timeouts when using Microsoft iSCSI software initiator for more details on timeout configuration). An I/O could take a longer time to execute if the filer is slow to respond to I/Os. Usually the Storage stack will retry the failed I/O's for a couple of times before giving up. The I/O retries will be transparent to applications above the storage stack. So as long as your application runs fine such occasional error message should not be cause for concern.
- I am seeing the same error messages as mentioned in question #1. However my applications do not run fine and they crash. What should I do?
This should be addressed by a combination of 2 factors.
- Microsoft Initiator Specific Tunable:
There is a Microsoft Initiator specific tunable, SrbTimeoutDelta, which can be tuned to handle filer latencies. See KB articles: Host-side SCSI Timeouts when using Microsoft iSCSI Software Initiator and Host-side SCSI Timeouts when using iSCSI for a description of this tunable in detail and steps on how to tune it. - Performance Tuning:
The techniques used to address filer performance issues with other protocols (NFS, CIFS) apply to iSCSI as well. This should improve the filer latencies and should eliminate or reduce considerably the error messages. Refer ntapcs 16891 and follow the procedures discussed in the KB article.
- I am using iSCSI and I am seeing the error messages listed in question #1. I did not migrate from VLD. What needs to be done?
These error messages are logged when the Client side SCSI Storage stack tries to execute an I/O, and the I/O did not complete within a specified period of time (the default timeout value is 25 secs for Microsoft software initiator).
An I/O could take a longer time to execute if the filer is slow to respond to I/Os. Usually the Storage stack will retry the failed I/O's for a couple of times before giving up. The I/O retries will be transparent to applications above the storage stack. So as long as your application runs fine such occasional error message should not be cause for concern. However, if your applications are failing then you can take the following remedial measures:- Microsoft Initiator Specific Tunable:
There is a Microsoft Initiator specific tunable, SrbTimeoutDelta, which can be tuned to handle filer latencies. See KB: Host-side SCSI Timeouts when using Microsoft iSCSI Software Initiator for a description of this tunable in detail and steps on how to tune it. - Performance Tuning:
The techniques used to address filer performance issues with other protocols (NFS, CIFS) apply to iSCSI as well. This should improve the filer latencies and should eliminate or reduce considerably the error messages. Follow the procedures discussed in the KB: Host-side SCSI Timeouts when using iSCSI.
- Microsoft Initiator Specific Tunable:
Additional Information
additionalInformation_text