Solved!
Go to SolutionFailed to "rescan drv_cfg ".
We created volumes of ScaleIO, and mapped to Windows server where SDC has already been installed.
Then, we run " drv_cfg --rescan " to detect new volumes, however, the rescan failed with an error message:
" Failed to open kernel device" .
Please show us the cause and workaround of this error.
Also, could you let us know the way to see the created volume from the server?
The procedures that we have been done are as below:
1. Create volume
cli --add_volume --protection_domain_name default --storage_pool_name default --size_gb 16 --volume_name svol1
2. Map
cli --map_volume_to_sdc --volume_name svol1 --sdc_ip
3. drv_cfg Rescan
C:\Program Files\EMC\scaleio\sdc\bin>drv_cfg --rescan
Calling kernel module to refresh MDM configuration information
Failed to open kernel device
Thank you for your quick responce.
The Volume is not displayed in Disk Management window on we Windows OS.
1. cli --query_all_sdc command
C:\Program Files\EMC\scaleio\mdm\bin>cli --query_all_sdc
MDM restricted SDC mode: Enabled
Query all SDC returned 3 SDC nodes.
SDC ID: 369ed70200000000 Name: sdc1 IP: State: Disconnected GUID: 00000000-0000-0000-0000-000000000000 Approved: yes
Read band 0 IOPS 0 Bytes per-second
Write band 0 IOPS 0 Bytes per-second
SDC ID: 369ed70300000001 Name: sdc2 IP: State: Disconnected GUID: 00000000-0000-0000-0000-000000000000 Approved: yes
Read band 0 IOPS 0 Bytes per-second
Write band 0 IOPS 0 Bytes per-second
SDC ID: 369ed70400000002 Name: sdc3 IP: State: Disconnected GUID: 00000000-0000-0000-0000-000000000000 Approved: yes
Read band 0 IOPS 0 Bytes per-second
Write band 0 IOPS 0 Bytes per-second
2. cli --query_all_volumes command
C:\Program Files\EMC\scaleio\mdm\bin>cli --query_all_volumes
Query-all-volumes returned 2 volumes
Protection Domain 8bbe0b0100000000 Name: default
Storage Pool c04242c300000000 Name: default
Volume ID: cc17946f00000000 Name: svol1 Size: 16.0 GB (16384 MB) Mapped to 1 SDC Thick-provisioned
Volume ID: cc17bb8000000001 Name: svol2 Size: 16.0 GB (16384 MB) Mapped to 1 SDC Thick-provisioned
3. showevents.py (can be found in mdm bin folder)
We attached the file "showevents.py".
1 Attachment
showevents.py
showevents.py
tomer__engineer
155 Posts
2283
1
Posted July 28th, 2015 01:00
Could it be that the volume already appears on your Windows OS? it will appear in Disk Management window as a new Disk (e.g. if you have Disk1-5, it will appear Disk 6 with the size that you gave the volume, in your case 16GB)
Please send the output of:
1. cli --query_all_sdc command
2. cli --query_all_volumes command
3. showevents.py (can be found in mdm bin folder)
tomer__engineer
155 Posts
2283
1
Posted July 28th, 2015 06:00
I see you are using the restricted SDC mode and it's enabled.
If you wish to use it you need to perform an --add_sdc command (press enter to see the usage) for each sdc in order to enable it to connect with the MDM.
Did you perform this command? all SDCs are currently disconnected and that is why none of then can "see" the volumes you are trying to map.
Once you add all the SDCs, run the --query_all_sdc command again and you should see all of the connected, I think they will also see the volumes, if that does happen just unmap the volume and map it back again.
Restricted SDC mode enabled is not the default - did you enable it on purpose?
Thank you for your response.
Now, we understood that "drv_cfg --rescan" error (mapped volumes cannot see from the server) was caused by that the patch "KB3033929" had not been applied to Windows server where SDC was installed.
After installing the patch, we mapped the volumes.Then, we could see the volumes from the server.
This issue has been solved.
Thank you so much for your help.
tomer__engineer
155 Posts
2283
0
Posted August 2nd, 2015 23:00
Good to hear. I think this issue can be closed.