The pt3agent is up and running, but in the .err and .debug, it looks like the agent is repeatedly trying to start. Why?
This happens when the command used to look for already running agents does not find the already running agent. By default, the command used to list potential agents is:
ps -C _progres -f | grep "p util/pt3agent”That should return a list of running pt3agent processes with the PID in column 2. Like this:
protop> ps -C _progres -f | grep pt3agent
oedba 73307 1 2 Jun23 ? 02:19:21 /usr/dlc124/dlc/bin/_progres -b -p util/pt3agent.p -param s2k /db/s2k.124/s2k124 ...
If the PID is not in column 2, the command can be changed via the GETAGENTLIST environment variable in bin/localenv.
If you don't have a bin/localenv, create one by copying bin/localenv.x to bin/localenv. Then, uncomment the line containing GETAGENTLIST and alter the command to show the PID in the second column.
We also check to see if the process-id in the flg file exists by using:
ps -p &1 -f
Where &1 is the PID that we are looking for… like so:
protop> ps -p 73307 -f
UID PID PPID C STIME TTY TIME CMD
tom 73307 1 2 Jun23 ? 02:19:21 /usr/dlc124/dlc/bin/_progres -b -p util/pt3agent.p -param s2k /db/s2k.124/s2k124 ...
The GETPID environment variable can override that.
If these commands do not behave, it might be because you are using something other than “sh” or "bash" for the shell.
Otherwise...
If the preceding doesn't clear up the issue, set a higher debug level for the dbmonitor process and review the $PROTOP/log/dbmonitor.log for references to chkDB and surrounding lines for interesting messages.
If all else fails, contact us via the chatbot on this page or use the link at the top. We will get back to you as soon as possible.