| Author |
Message
|
| kalam475 |
 Posted: Tue Feb 28, 2017 12:48 am Post subject: Disaster Recovery Setup |
 |
|
Acolyte
Joined: 16 Jan 2015
Posts: 63
|
Hi,
We have a three Multi-instance queue managers (CLUSQM1 & CLUSQM2 & CLUSQM3 )which are participating in the Clustering. CLUSQM1 and CLUSQM3 are full repositories and CLUSQM2 is half repository. CLUSQM1 and CLUSQM2 are hosting the Cluster queues and CLUSQM3 is there to maintain the CLUSTER and doesn't host any queues.
Now we want to set up a DR for the same. Which we are thinking of creating a backup queue manager for each queue manager. In-order to create the backup queue manager it should resemble the actual queue manager such as
1.queue manager name,
2.the logging type,
3.and the log file size.
apart from these is there any other extra consideration to be taken since these queue managers are participating in the clusters. If so any suggestion would be good. Knowledge center doesn't provide any specific details for creating a backup queue manager participating in cluster.
If no need of changing any thing how cluster queue effect in this case and Cluster channels needs to be created or not.
Also apart from backup queue manager is there any alternative for Disaster recovery. Please spare my ignorace here. Thanks |
|
| Back to top |
|
 |
| smdavies99 |
| Posted: Tue Feb 28, 2017 3:19 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
You do have access to the scripts that were used to create the QMGRS in the first place?
Are all the changes (like adding Queues, Channels etc) since creation scripted and in Source control?
If not then perhaps now is a good time to create them and you know actually use them to create your DR environment?
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| kalam475 |
| Posted: Tue Feb 28, 2017 8:41 am Post subject: |
|
|
Acolyte
Joined: 16 Jan 2015
Posts: 63
|
I do have all the commands how i created the queeue manager. but doesn't have the commands of how i created the queues. but if i run rcdmqimg before taking the backup doesn't it take the image of the queue we have created. let us say before moving the log files i created a queue and after moving it to DR and replay doesn't it create the queue I guess it will.
Any how my question was since my QMGRs are participating in the cluster do i have to take any precatuions or backup queue manager will not support the qmgrs participating in the clusters or there are cluster channels are there are they gonna effect my base queue manager?
Thanks for help:) |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Feb 28, 2017 9:06 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| kalam475 wrote: |
| I do have all the commands how i created the queeue manager. but doesn't have the commands of how i created the queues. |
All is not lost. You can generate them with the dmpmqcfg command.
Do some experimentation. your DR site is perfect for that.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| kalam475 |
| Posted: Tue Feb 28, 2017 11:10 pm Post subject: |
|
|
Acolyte
Joined: 16 Jan 2015
Posts: 63
|
Thanks for reply. I do have another question.
Let us suppose say I have created a queue manager with Linear logging of primary files of arround 200 so I have 200 log files available in my mount point /nfsshare/QMGRSlogs/CLUSQM1 and data files are at /nfsshare/QMGRSdata/CLUSQM1. In order to create a backup queue manager in my DR I use the same commands to create it.
Now for the first time I will stop my actual qmgr and copy the data and log folders to DR and replay the logs. now everything works fine. Once my qmgr in production starts and start filling up how do i keep both in sync.
1st option :
use some kind of replication (synchronous) and hope that both are in sync and when disaster strike i reply the logs and start the queue manager. Does this works? and simple.
2nd option:
At regular intervals I will run rcdmqimg command and take the logs (n-1) of media log and replace them in DR. run strmqm -r <QMGR>. when actual disaster happens copy all the latest log files and activate the qmgr. In this scenario there is chance of loosing the messages or duplicating the messages.
IF I go with 1st option any is still there a chance of loosing messages. Is there a possible scenario where we can't loose messages or duplicate messages. Any technote or redbook suggested is very helpful.
 |
|
| Back to top |
|
|
| mqjeff |
| Posted: Wed Mar 01, 2017 4:41 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
Ask your storage admins/dbas how this problem is solved for databases.
_________________
chmod -R ugo-wx / |
|
| Back to top |
|
|
| gbaddeley |
| Posted: Wed Mar 01, 2017 7:48 pm Post subject: |
|
|
Jedi Knight
Joined: 25 Mar 2003
Posts: 2538
Location: Melbourne, Australia
|
| Quote: |
| IF I go with 1st option any is still there a chance of loosing messages. Is there a possible scenario where we can't loose messages or duplicate messages. Any technote or redbook suggested is very helpful. |
In a true DR scenario (where not just MQ is affected), it is normal to hand over the DR qmgr in an empty state. It is the responsibility of apps to reconcile / reprocess any missing or duplicate transactions / messages.
This avoids the issue of trust in the MQ recovery logs, where there is a chance that a message may or may not have been logged at the point of disaster.
_________________
Glenn |
|
| Back to top |
|
|
|
|
