*Terracotta used and 10.7 level:
We restarted our terracotta server but it’s not starting up after shutdown of server.
We are getting below error in wrapper logs.
INFO | jvm 1 | 2024/03/24 22:45:22 | Fatal Terracotta startup exception:
INFO | jvm 1 | 2024/03/24 22:45:22 |
INFO | jvm 1 | 2024/03/24 22:45:22 |
INFO | jvm 1 | 2024/03/24 22:45:22 | *******************************************************************************
INFO | jvm 1 | 2024/03/24 22:45:22 | You have specified server name ‘serverName’ which does not exist in the specified tc-config file.
INFO | jvm 1 | 2024/03/24 22:45:22 |
INFO | jvm 1 | 2024/03/24 22:45:22 | Please check your settings and try again.
The server was working fine 2 hours back.
We have checked the /etc/hosts file the hostname and ip are mapped correctly.
Also check that parameter.6 property is set to %Hostname% and changed it to server hostname but didn’t work.
Any support is highly appreciteated, it’s block multiple on going projects.
Question related to a preprod instance?
Do you have your TC servers defined in the tc-config file?
How are you starting/ restarting the servers?
It seems the server name is not in the tc-config file.
Hi Wayne,
Thank you for your response, the tc-config file has server name TCCEServer1 but the error shows server name as different.
The terracotta server came up after starting with command nohup ./start-tc-server.sh -f ~/tc-config.xml -n TCCEServer1 &.
Generally we start from startup.sh under server/wrapper/bin and it was working fine.
This time startup.sh didn’t work so started with start-tc-server script.
I think there are some file which we need to delete from sever-data which cause this issue, if you are aware of those file can you file share any details.
Thanks
Depending on the startup script you use to execute the process terracotta uses a different tc-config.xml file. All your servers should be defined in this tc-config.xml file. According to the log you shared, a node in your cluster is not defined in tc-config.xml file is trying to register itself to the cluster, this cluster is managed by the first terracotta server you start. So this is the first tc-config.xml you should validate. It should have the whole cluster information and the rest of the cluster nodes should also have the same array definition.
If tc-config.xml doesn’t have the whole cluster information, that will cause a split brain. Depending on the order you start the processes, it may or may not shut down during startup process, so process not shutting itself down is not an indicator it was working properly. You should verify your tc-config.xml file, mainly the server array you have.
Hi Engin,
We suspect that its not picking up the correct file now, we didn’t do any changes to our configs or script but its picking the hostname in correctly.
When we start using -f option with -n instance name it works fine.
Thanks
Go to /opt/softwareag/Terracotta/server/wrapper/conf and view wrapper-tsa.conf file. Configure the tc-config.xml file you see there. If the configuration file looks good, start terracotta by running /opt/softwareag/Terracotta/server/wrapper/bin/startup.sh or by starting the deamon. Replace /opt/softwareag/ with your softwareag installation directory if installed to a different location other then default. nohup is not required for this process, it will keep your process alive even if your session terminates. Repeat this for every node you have.
Hi Engin,
Thanks for the information. I just checked and found that the wrapper-tsa.conf file was taking server name from the value configured as a parameter.6 here.
We’ll change the servername to correct one and see if this helps.
Thanks