Unsolved
This post is more than 5 years old
6 Posts
0
3821
VMAX, SRDF configuration , FCIP with limited bandwidth
In our current configuration we have an FC over IP connection between our two datacentres with a limited bandwidth (200 Mb /sec). We have 2 DMXs and SRDF/A configured. The bandwidth was sufficient to cover the SRDF/A traffic between our 2 DMXs.
Recently, we installed two new VMAXs, one in each datacentre and I configured SRDF groups to establish an RDF connection between our two VMAXs then I configured one pair of devices (mode acp disk) to test the synchronization process.. Everything went well until I started the synchronization. Once the synchro was running, the command "symrdf -g srdf_test query" was failing with the message : The Symmetrix configuration could not be loaded for a remotely-attached Symmetrix. and I realized that I was not able to run any symcli RDF command , all were hanging !!
The SE version installed is 7.1.1.0
Based on that , here are my questions:
- Is there any parameter which could be set on the VMAX to specify that we have a limited bandwidth ?
- Should I ask EMC CS to activate SRDF RF flow control (emc152051) ?
- for my SRDF/A configuration, should I setup DSE ? if yes, what are the EMC recommendations ?
- Which other settings I should consider ?
I would appreciate if you could give me some guidelines or advices.
Thanks
dynamox
1 Rookie
1 Rookie
•
20.4K Posts
0
July 20th, 2010 07:00
symrdf -sid ping
comes back "Successful" ?
pat4me
6 Posts
0
July 20th, 2010 07:00
yes ping works :
symrdf ping -sid 61
Successfully pinged (Remotely) Symmetrix ID:
pat4me
6 Posts
0
July 20th, 2010 07:00
I don't think I have an issue with the daemons, I tried to run this query from 2 different hosts, one connected to R1 box and the other one connected to R2 box, same result.
As soon as I stop the synchro , I can run my command without any problem. It's why I was thinking it could be related to my bandwidth , in other words when the synch is running it uses the whole bandwidth and symrdf command can't reach the remote VMAX !
AJK3
108 Posts
1
July 20th, 2010 07:00
I dont think its entirely a bandwidth issue. Its issue with SYMCLI or SE host that you are running the commands.
Is this intermittent or continuous?
I had seen this kind of issue earlier some time and got this resolved by restarting the local daemons or restarting the host.
I will wait for others inputs to see if there is any thing else involved in this set up. I am eager to check if it applies to my environment as well.
dynamox
1 Rookie
1 Rookie
•
20.4K Posts
0
July 20th, 2010 14:00
i would open a ticket with support, curious to know what the problem is and what they recommend (besides get a bigger pipe).
beetgj
16 Posts
1
July 27th, 2010 07:00
request a SVC qualifier from your EMC representative to qualify your SRDF environment. Thier is a lot of variables and options that needs to be considered, you can also talk to a EMC solutions architect. SRDFA link bandwidth get calculated according to the cache in the box and the amount of change tracks that needs to be shipped, this will also determine the cycle time for shipment....everytime you upgrade capacity this calculation needs to be revisited....DSE should not be an option to balance the environment and should only be used in a balance environment...DSE are their for keeping SRDFA up and running when the links go down for a limited period of time without saturating cache, meaning it destage to disk and don't use all your write slots in cache...so your delta sets don't fill up cache but sits on disk while it waits for the links to recover