Error message with DR

A

Anne-Marie

New Member
Hi

I am a still new to this and now a DR is being looked at, the following error is coming up, any ideas or suggestions would be greatly appreciated.



We currently have a Novell PlateSpin Forge Appliance replicating (P2V) incrementally both coins Linux Red hat servers, this is done at block level using a custom kernel based BBT driver.

In Tests we can easily achieve getting the Red Hat Machines back in a consistent state including all of the local and ISCSI attached file systems. It does this by converting the externally connected ISCSI LUN’s into virtual disks.

Because the Red Hat host is configured with named volumes, when they are failed over the file systems auto mount which works very well. The issue is that we cannot get the coins system started we get the following error:

Connecting to Progress AdminServer using rmi://localhost:20931/Chimera (8280)
Searching for oalive (8288)
Connecting to oalive (8276)
Unable to start oalive (8297)
Running /u/coins/live/bin/watchdog_coins.sh
17:39:24 WDOG There is no server for database /u/coins/live/data/db/coins. (1423)
17:39:24 WDOG ** This process terminated with exit code 8. (8619)
Running /u/coins/live/bin/apw_coins.sh
17:39:24 APW There is no server for database /u/coins/live/data/db/coins. (1423)
17:39:24 APW ** This process terminated with exit code 8. (8619)
Running /u/coins/live/bin/biw_coins.sh
17:39:24 BIW There is no server for database /u/coins/live/data/db/coins. (1423)
17:39:24 BIW ** This process terminated with exit code 2. (8619)
Running /u/coins/learn/bin/broker-start_coins.sh
OpenEdge Release 10.2B02 as of Wed Sep 1 19:14:44 EDT 2010
17:39:24 BROKER ** The database /u/coins/learn/data/db/coins is in use in multi-user mode. (276)
17:39:24 BROKER ** This process terminated with exit code 2. (8619)
OpenEdge Release 10.2B02 as of Wed Sep 1 19:14:44 EDT 2010

Thanks
Anne-Marie
 
Looks like you have two environments "live" and "learn" and you are mixing and matching the startup scripts from these environment.

The database must be running before you are able to start a watch dog or asynchronous page writers against it. The same is true for the AppServer.


Heavy Regards, RealHeavyDude.
 
Hi RHD.

Would using the DBMAN command be the best way of checking if its running.

ie dbman -host -port -database -query

Thanks for your help.
Anne-Marie
 
IMHO - it does not really matter how you start the database

  • using dbman utilizing the AdminServer framework in a "managed" way
  • directly using proserve in an "unmanaged" way
it must be running before you attempt to start additional database processes like the watch dog or the asynchronous page writers. The same is true for the AppServer: Out of experience I can tell that when an AppServer fails to start, 99% of the time, it is either that the database is down or the connection parameters to the database are not correct.

But, yes you can use dbman to check whether the database is running and there are other ways too depending on how your system is set up.

You are saying that you can't start the coins system. Personally I don't know anything about a coins system. How do you start the system? Do you use a script to start the system? If that's the case what does the script do?

From the error messages and what you are telling I can just make vague guesses. Plus I am still under the impression that you try to start the "live" system - which fails and as a response try to start the "learn" database - which fails too as the "learn" database is already running.

Heavy Regards, RealHeavyDude.
 
Hi

Thanks for the respsonse, however, it appears that its an incremental backup that was being attempted to restore, although I had mentioned that these cant be used. Until there has been an attempt with a full-back up I will leave this thread as it is and if the same error comes up then will revisit.

Many thanks for your help and time.

Anne-Marie
 
IMHO - it seems to me that the disaster recovery strategy and database administration need a serious review. Don't get me wrong, but out of experience I can tell you that - if you lose data or can't fulfill your service level agreements then somebody will be held responsible for that. If I were you I would not take that responsibility until I got proper training.

Heavy Regards, RealHeavyDude.
 
I agree, which is why I am currently awaiting a call back to arrange for someone at Progress to assist with this.

Anne-Marie
 
You must log in or register to reply here.
Back
Top