Unsolved
This post is more than 5 years old
2 Intern
•
48 Posts
0
1062
February 21st, 2014 07:00
Issue migrating Win2008 hosts to VMAX
We're seeing an odd problem when trying to migrate Win2008 hosts to VMAX. We're using Open Replicator so we ask the server admin to send us the output of a tool called "inq" provided by EMC. Running this tool with no command line options produces a report showing Disks and capacity. We need this to create the open replicator config file.
I've attached good and bad output files here but essentially, with some Win2008 hosts, the inq reports the capacity of all the current SAN storage on the host as "FAILED". The host reports no problem with the storage nad volumes are mounted and Apps are running on the storage with no indication of an issue.
We could use other means to create the open replicator config and we did the first time we saw this issue. The storage migration went well but the host had subsequent issue with the powerpath installation. The installer for 5.7 would die with no error message and eventually we were able to install 5.5 but the host wouldn't boot afterwards. Recovery was painful.
Has anyone seen this issue with the inq tool and Win2008 servers. We've seen it with both clusters and stand alone hosts.
chirswilson
34 Posts
0
February 21st, 2014 07:00
are you running it from a CMD prompt that was run as admin? (right click "run as admin")
vic_engle
2 Intern
•
48 Posts
0
February 21st, 2014 08:00
Yes, we start the CMD window with "Run as Administrator" and confirm the top of the CMD window says Administrator.. We ran into that early on.
dynamox
11 Legend
•
20.4K Posts
•
87.4K Points
0
February 21st, 2014 11:00
Issue INQ shows dashes "------" in the capacity field for some devices.
INQ shows "NA" in the Capacity field. INQ may show "?" for the Meta count field.
/dev/rhdisk13 M(?) EMC SYMMETRIX 5568 39063000 N/A
INQ shows "failed" in the capacity column.
Environment System: IBM RS/6000
OS: IBM AIX
Cause When INQ is run, it goes out over every I/O Bus to every disk and tries to read the vendor information off the disk drive itself. If INQ cannot get to a disk, due to locks by another host or very high I/O, INQ will get the information needed for that disk from the ODM. The only piece of data INQ cannot get from the ODM is the disk capacity. The INQ utility itself adds the dashes in the CAP:(MB) field for that device. In this environment the dashes are nothing to worry about. Also see INQ shows FAILED....
Cases have also been seen where orphan (hung/stray/ghost/runaway) processes, like EMC Workload Analyzer, are placing the lock on the device. In these cases, when the application (or hung process) was stopped (or killed), INQ was able to be run repeatedly without the dashes ("------" ).
Volume group had not been varied off before unmapping device from front end. The SCSI reservation will still be help by this director and device will be inaccessible.
Another AIX host with access to the same devices, has them currently varied-on in a volume group, thus placing a reservation on the devices
Resolution Use the "ps -ef" command to look for other processes that may be using the devices showing up with dashes in the INQ.
Customer may be able to release reservation with Solutions Enabler if the device has not been unmapped from front end. See solution 28651 ("Can solutions Enabler be used to clear SCSI reservations").
chirswilson
34 Posts
0
February 21st, 2014 11:00
Seems similar to this :- https://support.emc.com/kb/46783
INQ shows dashes "------" or "Failed" in the capacity field for some devices.