This post is more than 5 years old
9 Posts
0
3695
ScaleIO Cluster went down and after the MDM's came back both are in Secondary state How do I fix this
I have a ScaleIO 1.32.5 cluster ,both the ESXi host which had PMDM & SMDM went down together.After they came back up both are showing as secondary MDM's.How can i fix this any advice will be great ?
SanjeevMalhotra
138 Posts
0
August 20th, 2016 14:00
Assisted on SR opened by Rishi.
SysEng777
25 Posts
0
August 19th, 2016 14:00
Have you tried running the command on the MDM to swap between primary and secondary? I was wondering if your just having a display issue and by issuing the command it might correct that. Page 324 of the user guide.
scli --switch_mdm_ownership (--new_master_mdm_id |
--new_master_mdm_ip [--new_master_mdm_port
]| --new_master_mdm_name )
SanjeevMalhotra
138 Posts
0
August 20th, 2016 07:00
You may try to restart the MDM service on both the MDMs one by one.
Stop service:
/opt/emc/scaleio/mdm/bin/delete_service.sh
Start service:
/opt/emc/scaleio/mdm/bin/create_service.sh
Thereafter again check the output of "scli --query_cluster"
SanjeevMalhotra
138 Posts
0
August 20th, 2016 07:00
Can you please post the output of "scli --query_cluster" here. You may mask the IP addresses in the output before posting them here.
rishizalani
9 Posts
0
August 20th, 2016 09:00
Can't use the command since both are secondary
SanjeevMalhotra
138 Posts
0
August 20th, 2016 09:00
Check https://support.emc.com/kb/482118
rishizalani
9 Posts
0
August 20th, 2016 09:00
This is the error I get after service restart :
scli --query_cluster --use_nonsecure_communication
Error: MDM failed command. Status: The command used for Tie-Breaker is not allowed
I guess now all the three are showing as Tie Breaker
SanjeevMalhotra
138 Posts
0
August 20th, 2016 09:00
Is the the same error when you run "scli --query_cluster" on all the 3 MDMs?
rishizalani
9 Posts
0
August 20th, 2016 09:00
Since both MDM's are in secondary state you can't issue scli commands
SanjeevMalhotra
138 Posts
0
August 20th, 2016 09:00
I will suggest to collect get-info from all the MDMs and open a Service Request with Tech Support and we can investigate.
rishizalani
9 Posts
0
August 20th, 2016 14:00
Sanjeev thanks for your help .Keep up the good work.
SysEng777
25 Posts
1
August 22nd, 2016 08:00
Seems like it would be more helpful to the community if the answer was posted vs an SR number.