This post is more than 5 years old
1 Rookie
•
12 Posts
0
12201
April 13th, 2018 23:00
Live Volume - No fail-over
Hi,
as described in SC-Series-Synchronous-Replication-and-Live-Volume-Dell-EMC-2017 i´ve configured the live volumes with ha + auto fail-over on two mid-range SCvXXXX.
This the first time for me using the live volume feature.
- I´ve zoned the esxi hosts 6.0 to every vWWN (source and target compellent).
- i deployed two datacollectors/dsm (prim. and sek.)
- I configured the volumes for auto-failover and ha, no fail-back
Everything showed "up and running". My test was now to pull the target-cables (fc) of the prim. system. The test-esx lost the connection to the lun. What was going wrong?
- Do i have to map the target lun (named "lv of lun") also to the esx-hosts? While the system works in normal state before?
Regards
Max
No Events found!
DELL-Bob Mi
2 Intern
•
230 Posts
1
April 14th, 2018 10:00
I see you have deployed 2 data collectors, but you do not mention if you deployed a tiebreaker. See page 52 of the Synchronous Replication and Live Volume Best Practices. Also, it is recommended to use Fixed for the MPIO in VSphere per page 45 of the guide.
http://en.community.dell.com/techcenter/extras/m/white_papers/20439310/
DELL-Bob Mi
2 Intern
•
230 Posts
1
April 15th, 2018 16:00
Was this test even realistic or does disconnecting the front cabling not bring a fail-over to the secondary side?
If a network connection is lost then no auto failover will occur. This was shown in the Dell Storage SC Series 6.7 Live Volume Automatic Failover video
http://en.community.dell.com/techcenter/extras/m/white_papers/20441883
What do you suggest for testing to show auto-failover to a customer (without "stressing" the hardware)?
In that video they demonstrated by rebooting both controllers on the Primary Site. I am not sure that is an option for you.
What about server objects under "servers" of the secondary system?
Do I even have to create the servers (initiators) there?
Yes, for proper testing they should be configured for connection
Compellentuser99
1 Rookie
•
12 Posts
0
April 15th, 2018 00:00
Hello Bob,
thank you for the document and the note with the "tie breaker".
I forgot to mention the tie-breaker and of course installed it.
Please note that my test was to remove the front-end by disconnecting the front wiring of the primary site. Included in front cabling: Target ports for servers (ESXi hosts) and for replication.
Tiebreaker was displayed as online. All other states were green according to primary dsm/dc. Essential questions for my part are now:
Best regards
Compellentuser99
1 Rookie
•
12 Posts
0
April 15th, 2018 21:00
Thank you for your advices.
Compellentuser99
1 Rookie
•
12 Posts
0
April 20th, 2018 05:00
Hi,
I just wanted to let you know that everything has worked - smoothly.
regards
Ibrahim Akram
2 Posts
0
May 9th, 2018 02:00
Hi,
I am facing the same problem,
could you advice please how you fixed it.
i dont know if the tiebreaker is configured properly or not.
its appear like checked green but the status of the local tiebreaker is down (primary and secondary nodes)
is there any configuration should i do to bring them up??
THANKS.
anhnh1977
1 Rookie
•
19 Posts
0
November 10th, 2018 07:00
Hi,
We have the same issue as the local tiebreaker is down (primary and secondary nodes).
Is there any document which show how to setup/configure Tiebreaker???
Rgds,
DELL-Bob Mi
2 Intern
•
230 Posts
0
November 10th, 2018 09:00
I posted this on another Forum post:
The best article on Tiebreaker is on page 577 of the Storage Manager 2018 R1 Administrator’s Guide:
https://topics-cdn.dell.com/pdf/storage-sc2000_administrator-guide3_en-us.pdf
The Tiebreaker should be located at a third autonomous site which can monitor both the Primary and DR sites. It should be configured when setting up Live Volume with Auto Failover.