rasynchronous san replication configuration dr box build

rammaghenthar
rammaghenthar used Ask the Experts™
on
hi
         currently we are building  disaster recovery lpars for existing live systems. This task involves some file systems to be configure  asynchronous san disk replication with live systems. What points i need to consider while implementing asynchronous san replication?  in single vg may i configure only some list of filesystems as part of replication?
Comment
Watch Question

Do more with

Expert Office
EXPERT OFFICE® is a registered trademark of EXPERTS EXCHANGE®
Most Valuable Expert 2013
Top Expert 2013
Commented:
Hi,
there is no real SAN replication feature built into AIX.
Do you mean something like IBM's GlobalMirror or maybe SRDF from EMC?
These tools are not aware of logical volumes, however (at least afaik).
Which "disaster" are you planning for? Are the DR LPARs in a remote location? If not, you could consider taking over the VGs manually in the case of a disaster concerning the servers (not the SAN box(es)), just like HACMP does. You only need to attach the concerned LUNs to both LPARs, to do "importvg" if necessary.
As said, the above can be used to recover from server outages, not from an outage of the SAN box(es).
To provide for recovery from SAN box failures you could of course also go the "pedestrian" way and replicate using tools like rsync.
Anyway, please give more detail information!
 wmp
Commented:
do you mean replicating between sites? yes
there are two ways:
1- using copy services
2- using LVM mirroring
it could be... metro or global.

I would check the following link
http://www.ibmsystemsmag.com/aix/februarymarch08/focusonstorage/19296p1.aspx
http://www.ibmsystemsmag.com/aix/enewsletterexclusive/6757p1.aspx


"Metro Mirror is generally considered a campus-level solution, where the systems are located in fairly close proximity, such as within the same city. However, the distance supported will vary based on the write intensity of the application and the network being used. In general, with adequate resources, most customers find up to a 50-kilometer distance acceptable with some customers implementing up to a 300-kilometer distance. "With Global Mirror, the target site may trail the production site by a few seconds. The frequency of creating a consistency group is a tunable parameter, and you?ll need to balance your recovery point objective with the performance impact of creating a consistency group. Many customers find a three- to five-second consistency group achievable (i.e., in a disaster, you?d lose the last three to five seconds of data)."
Reference: http://www.ibmsystemsmag.com/ibmi/march09/features/24370p1.aspx
Commented:
what about HACMP/XD GLVM over a WAN lines (T1,E1 or E3)....for failover and keeping data on both sites sync'd....HACMP/XD can be setup using copy services or LVM mirroring the latter is called cross-site LVM mirroringin case of copy services, HACMP will send command to disk subsystems to make them online or offline for application

PMI ACP® Project Management

Prepare for the PMI Agile Certified Practitioner (PMI-ACP)® exam, which formally recognizes your knowledge of agile principles and your skill with agile techniques.

Author

Commented:
i am talking about  3 site storage technology primary,sec,teritary site. Project team provided some mount points be configured with asynchronous with live . What information do i need to provide to storage?
Commented:
"asynchronous san replication?  in single vg may i configure only some list of filesystems as part of replication?"  
No it wont work in that way, if you consider hardware replication you have to replicate the complete LUN!

in aproject I have implemented 3 sites replication (Hardware SAN replication) using IBM technology DS8100 SAN storage. I have SAN to SAN  replication synchronous over Dark-fiber (distance=5Km) with zero lost data between Site A and Site B (Metro Mirror), and asynchronous replication over IP/VPN MPLS Layer3 (no Internet traffic) with 35Mbps (distance=150Km) between Site B and Site C (Metro Global Mirror), and MGM metro global mirror between Site A + Site C to get the meta data between the sites. .. is this what  you are planning to do?

Author

Commented:
waiting for project decision on ths task. We have updated required prerequisites
Commented:
For using 3 sites replication A-(MM)-B-(GM)-C in IBM solution; please be aware since the Global Mirror is most exposed due to the bandwidth constraints is most important to protect the Global Mirror target data with special respect of avoiding another initial copy to the remote site by either accidentally removing the Global Copy relations or by site failure of the intermediate site.

This can be avoided by implementing Incremental Resynchronization. It requires to have another physical link from the local site A  to the remote site C. This has been implemented by routing the traffic from the local site A via the existing dark fiber connection to the intermediate site B where the access point of link to the remote site C resides. Incremental Resynchronization does not require additional bandwidth to the remote site C, because the local site is used mostly for meta data exchange or for data traffic alternatively to normal Global Mirror.

Author

Commented:
currently project is planning for application level replication waiting for more information

Author

Commented:
currently live environment data is sitting on hds storage to be moved to ibm svc
If the source and target storage are different, then you may have to consider doing OS based replication.
That means each VG will have the LUNs from one storage and they are mirrored with the LUNs from another storage. You know, what I'm talking about.

Author

Commented:
...

Do more with

Expert Office
Submit tech questions to Ask the Experts™ at any time to receive solutions, advice, and new ideas from leading industry professionals.

Start 7-Day Free Trial