Quickplace Admin

Replicamaker errors

ReplicaMaker errors

On the second server in a cluster you have a program that run's called replicamaker. It should be a program document and will look something like this
load qptool replicamaker -a -s server1
When you add a new quickplace to a cluster it goes on server1. This task ensures that server1 and server2 are in line. i.e. contain the same quickplaces
When this task runs you might see something like this in your log.nsf
Running Program document for qptool replicamaker -s -s server1
start|Config::ReadLtpaConfig()
End|Config::ReadLtpaConfig()
JVM: Virtual Machine initialized.
ReplicaMaker: Starting to create replica stubs between Server1 local
File truncated - file may have been damaged
File already Exists
ReplicaMaker: Created 0 replicas out of 1 on server Server1
ReplicaMaker: Created 0 replicas out of 1 on server local
ReplicaMaker: Finished created replica stubs between Server1 and local
qptool: Command finished: replicamaker

Now how great an error message is that. I have some 3,000 quickplaces between these two servers and two of them have problems.
What you need to do is to ask the replicamaker to give you more info next time it runs
On the admin console for the server you need to use the command Set config QuickPlaceStubMakerLogging=3 (No need to bounce anything on the server
Then next time ReplicaMaker runs the log will have the name of the quickplace in question
You can then issue the command Set config QuickPlaceStubMakerLogging=0 on the console and things will be back to normal (again not need to bounce anything on the server

When I find the databse giving the issue I can then fix it. I often see that one file on a server might be 100Meg and the file on the second server is 10k. This means the file on the second server got currupted the first time that the replicamaker tried to create the stub. To get around this I have gone into the OS and manualy deleted this stub of a stub from the second server. Then next time replicamaker runs it creates the stub and everything is working fine.