?
Solved

Sybase Replication Server migration from Solaris to Linux

Posted on 2012-08-24
9
Medium Priority
?
1,607 Views
Last Modified: 2012-11-07
Hi

We want to migrate Sybase Replication Server along with OpenSwitch  from Solaris to Linux platform. Is there any migration document/checklist available on the web?

thanks
CS
0
Comment
Question by:CHAN
  • 4
  • 4
9 Comments
 
LVL 24

Assisted Solution

by:Joe Woodhouse
Joe Woodhouse earned 624 total points
ID: 38332413
I'm not aware of any documents generally available that cover all of this in one place.

Are you moving just Replication Server? Not any of its primary or replicate ASEs? What about its own RSSD ASE, or are you using ERSSD?

If the answer is just RS and not any ASEs at all, then I think it's far less work to treat this as a disaster recovery rather than a migration. Install a new RS, and then recover/resync the replicated databases.

If for some reason that isn't an option, it gets a lot more complicated because unlike ASE, there's no RS equivalent to move stable queues from one platform to another. At a minimum you'd be doing this with quiesced replication, empty queues, and probably rebuilding the queues afterwards.

Summary: read the RS Troubleshooting Guide on what's involved in recovering a RS, it will help you plenty here. If you're moving ASEs as well, that gets more complicated as there are a lot of hoops to jump through to make cross-platform dump & load work correctly.

As for OpenSwitch... I suspect you're looking at a full reinstall there.
0
 

Author Comment

by:CHAN
ID: 38332713
thanks Joe.

We are moving just Replication Server and OpenSwitch components to Linux.  RSSD is running on ASE.  Since we cannot move stable queue to new hardware, can we install new RS on Linux and load RSSD from the dump and start repagent on active ASE to reestablish replication? does it work? Our setup is WarmStandby with repdefs.


thanks
CS
0
 
LVL 24

Expert Comment

by:Joe Woodhouse
ID: 38332770
It's a bit more complicated than that. Your RSSD would not match your (newly built) stable queues. Check the resources I suggested as they'll show you what you need o do to fully rebuild replication.

I think it may take longer but will be a lot easier and less risky to fully rebuild your entire replication system. It's good practice for your DR too.
0
New feature and membership benefit!

New feature! Upgrade and increase expert visibility of your issues with Priority Questions.

 

Author Comment

by:CHAN
ID: 38353029
Hi Joe

We need to migrate Sybase Replication Server from Sun m4000 to Sun T4-2 series server.
Currently the stable queue has been setup on file system and RSSD on raw device. This is my implementation plan :

1. Remove secondary trunc marker on primary database.
2. Shutdown original RSSD & REPSERVER on m4000
3. copy the whole sybase directory from m4000 to T4-2 server
4. copy the stable device directory from m4000 to T4-2 server
5. Setup new RSSD on T4-2 server
6. Load RSSD from dump taken on m4000
8. run rs_zeroltm
9. start the rep agent on primary

Can you please review this plan and confirm whether it works? I appreciate your help.

thanks
CS
0
 
LVL 24

Expert Comment

by:Joe Woodhouse
ID: 38353058
I don't have the Replication Server Troubleshooting Guide handy so I can't comment in detail. Some things don't look right, though.

If you remove the secondary truncation point, you need to put it back. When will you do that? If you aren't actually resyncing your primary database to the replicate, why do this at all?

Are you taking an outage in your PDS while you do all this? If not, your replicate will be out of sync.

Given you're not actually changing or moving the PDS or RDS, I'm not sure you need to play with truncation markers.

Will the new hardware have exactly the same hostname and IP address of the old? If not, you'll need to update interfaces files (there, and on all ASE boxes involved in this).
0
 

Author Comment

by:CHAN
ID: 38354861
thanks for your comments Joe.

point # 9 is to put the secondary truncation marker and restart the rep agent on PDS. The ip address of new server will be same as old so there is no change in interface file. I am not sure about the value in rs_marker from new RSSD so I decided to reset the rs_marker value.
0
 
LVL 24

Expert Comment

by:Joe Woodhouse
ID: 38356815
If you're disabling and then later re-enabling the replication truncation marker, you must either enforce a total outage and zero writes in the PDS, or resync PDB to RDB afterwards. Any changes will be lost otherwise.
0
 
LVL 1

Accepted Solution

by:
BenSlade earned 621 total points
ID: 38364432
I think you can do the following (assuming big/little endian architecture matches up for the old/new systems)

    Shutdown the repserver

    dump load the RSSD to a new ASE server

    Make sure the new ASE server has the same name as the old one (rep agents need to connect to it, etc)

    Copy the stable queue files to the host where the new repserver will run.   The stable queue device files (you said your stable queue devices are using files) must be accessible by the new repserver using the same directory path as the old repserver server.

    use md5sum to make sure md5 checksums are the same on the old and new stable queue device files

    Start up the new repserver using

When we've done this, we've started up the new repserver using alternate ports and where the repserver's interfaces file has no entries for any of the other production ASE servers, but an test entry for a special test ASE server.  We then modified the interfaces file for the one test ASE server that used to connect to the old repserver (and should connect to the new repserver) to make sure the new repserver is replicating properly.  

If you get errors, shutdown the new repserver and change the interfaces files to point everything back to the original system and restart the old repserver.

All feedback (especially by Joe W) is appreciated.

Ben
0
 

Author Comment

by:CHAN
ID: 38366386
thanks Ben. Much appreciate your comments.  I am not sure whether do we need to reset the secondary truncation marker when we connect Active database to new repserver. BTW we will lock all the users before replication server migration.
0

Featured Post

Prep for the ITIL® Foundation Certification Exam

December’s Course of the Month is now available! Enroll to learn ITIL® Foundation best practices for delivering IT services effectively and efficiently.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Often times it's very very easy to extend a volume on a Linux instance in AWS, but impossible to shrink it. I wanted to contribute to the experts-exchange community a way of providing a procedure that works on an AWS instance. It can also be used on…
Sometimes MS breaks things just for fun... In Access 2003, only the maximum allowable SQL string length could cause problems as you built a recordset. Now, when using string data in a WHERE clause, the 'identifier' maximum is 128 characters. So, …
Is your data getting by on basic protection measures? In today’s climate of debilitating malware and ransomware—like WannaCry—that may not be enough. You need to establish more than basics, like a recovery plan that protects both data and endpoints.…
Despite its rising prevalence in the business world, "the cloud" is still misunderstood. Some companies still believe common misconceptions about lack of security in cloud solutions and many misuses of cloud storage options still occur every day. …
Suggested Courses
Course of the Month14 days, 12 hours left to enroll

839 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question