Try to ssh into your RPA, and run command "get_box_states" then Enter.
For your replication requirement :
1. Yes, you can select specific LUN not every LUN will be replicated. You must create CG(Consistecy Group). CG contain Source LUN, replica LUN and Journal Volume. Journal volume size minimum is 5GB for non-distributed and 20GB for distributed CG but this is depend on your environment. Example you create one CG with 5 source LUN and 5 replication volume(replication volume at least as large as source volume) then typical estimate is 20% of the size all data disks but once again you must analyze your environment first then you can perform journal sizing.
2. I'm not sure about replication delay, RPA by default is asynchronous but you must understand how recoverpoint work first. You can set async or sync in policy menu and I dont think RPA support sync replication over WAN(just support CDP and CRR via Fibre Channel). I'm not sure about replication delay, but you can try compression level (not recommended if your environment have large data change and you just have minimum amount RPA in one cluster) or you can using System Optimized Lag policy. For default the Lag policy is disabled, this setting define RPO of CG and you can set manually in MB, GB, hour, minutes or seconds. For your info, Lag is measured when a write made by the server host reaches the local RPA and stops being measured when data reach target RPA / target journal volume. And remember about allow regulation, this is allow recoverpoint to control the acknowledgement of writes back to the host in case of bottlenecks or insufficient resources that otherwise prevent recoverpoint for replicating data.
And once again, you must analyze your environment first and then you can perform recoverpoint sizing plan. You must obtain the data change per day or per hour, you must determine RPO you need and after you can obtain that you can meet your expectation.
I have done some changes in the zoning on the SAN switch i.e added all the storage SP's and the RPA HBA ports in one zone only. Now the fiber connection issue has been solved.
We have a environment where the customer has 2 DC's and customer will be using DC 1 as production site for one month and then failing over to DC 2 for another month. This will be happening repeatedly every month. We have configured Recover point replication between 2 sites through CRR where DC 1 is primary site and the replication is configured accordingly keeping DC 1 as Production Site while configuring Consistency Groups. If after one month we will be doing the failover to another DC, what changes are to be done at the consistency group level so that DC 2 will become the production site for that month.
aldyth.m
24 Posts
0
March 28th, 2013 08:00
Hi Holyboy,
Try to ssh into your RPA, and run command "get_box_states" then Enter.
For your replication requirement :
1. Yes, you can select specific LUN not every LUN will be replicated. You must create CG(Consistecy Group). CG contain Source LUN, replica LUN and Journal Volume. Journal volume size minimum is 5GB for non-distributed and 20GB for distributed CG but this is depend on your environment. Example you create one CG with 5 source LUN and 5 replication volume(replication volume at least as large as source volume) then typical estimate is 20% of the size all data disks but once again you must analyze your environment first then you can perform journal sizing.
2. I'm not sure about replication delay, RPA by default is asynchronous but you must understand how recoverpoint work first. You can set async or sync in policy menu and I dont think RPA support sync replication over WAN(just support CDP and CRR via Fibre Channel). I'm not sure about replication delay, but you can try compression level (not recommended if your environment have large data change and you just have minimum amount RPA in one cluster) or you can using System Optimized Lag policy. For default the Lag policy is disabled, this setting define RPO of CG and you can set manually in MB, GB, hour, minutes or seconds. For your info, Lag is measured when a write made by the server host reaches the local RPA and stops being measured when data reach target RPA / target journal volume. And remember about allow regulation, this is allow recoverpoint to control the acknowledgement of writes back to the host in case of bottlenecks or insufficient resources that otherwise prevent recoverpoint for replicating data.
And once again, you must analyze your environment first and then you can perform recoverpoint sizing plan. You must obtain the data change per day or per hour, you must determine RPO you need and after you can obtain that you can meet your expectation.
pratikv1306
73 Posts
0
April 2nd, 2013 00:00
Hi Aldyth,
I have done some changes in the zoning on the SAN switch i.e added all the storage SP's and the RPA HBA ports in one zone only. Now the fiber connection issue has been solved.
Regards,
Pratik.
aldyth.m
24 Posts
0
April 4th, 2013 00:00
Great to hear that, so you can change the replication delay?..Just try and improve a bit, recoverpoint is fun
pratikv1306
73 Posts
0
April 11th, 2013 02:00
Hello All,
We have a environment where the customer has 2 DC's and customer will be using DC 1 as production site for one month and then failing over to DC 2 for another month. This will be happening repeatedly every month. We have configured Recover point replication between 2 sites through CRR where DC 1 is primary site and the replication is configured accordingly keeping DC 1 as Production Site while configuring Consistency Groups. If after one month we will be doing the failover to another DC, what changes are to be done at the consistency group level so that DC 2 will become the production site for that month.
Kindly suggest on the same.
Regards,
Pratik.