Learn how easy it is to sync an existing GitHub or Google Code repo to a SourceForge project! See Demo
before i start let me say that you guys did great work with the gt.m-product!
Okay, I need the ability to replicate our database-system to a secondary system. When i first began my checks, i've got an error about missing $gtm_repl_instance variable. But i can't find any information about that. To what do i have to set that variable? Is it just a directory or is it a databasefile?
By the way, our database-system is organized in different "UCIs" (We converted from MSM...). That means, our programs are switching between those "UCI" via $zgbldir and $zroutines while running. Each Globaldirectory has its own underlying database but uses also one union database for global access (%-Globals). Could that be a problem for the replication? Or better, how should I design the configuration for the replication?
Thanks a lot,
excuse me, right now i've found the thread from eggclub about Replication. Let me quick read that, maybe that's the solution for my problems.
Now, i've got start it until following error:
gtm-test:/brunie # $gtm_dist/mupip replicate -receiver -start -listenport=3000 -log=rec.txt
%GTM-E-RECVPOOLSETUP, Receive Pool setup error
%GTM-I-TEXT, Error with receive pool creation
%SYSTEM-E-ENO22, Invalid argument
What's going on now?
The replication processes use 64Mb shared memory segments (by default). This is generally larger than the default max shared segment size allowed by the linux kernel. To increase the max segment size in linux to 128Mb, edit the /etc/sysctl.conf file and add "kernel.shmmax = 134217728". The change will go into effect after the system is rebooted.
You should also start the passive replication process first using $gtm_dist/mupip replicate -source -start -passive