Re: Determining instance's oninits

On Mar 6, 9:34 am, theBP <th...@Usenet-News.Net> wrote: > red_val...@yahoo.com wrote: > > I have 3 informixservers running on a single host.  One of the > > instances has experienced an assertion failure and is now > > unresponsive, so I'd like to kill its oninit processes and attempt to > > restart it.  But running onstat -g ses only reports a PID of 0 for the > > oninit sessions.  How do I determine which oninit processes are > > associated with which instances? > > > Running IDS 9.40FC8 on HPUX11. > > onstat -g glo> > Establish the main CPU VP. > > Get an onstat -g stk all first, and an onstat -a for diagnostics. Thank you. Now that I've been able to identify the oninit processes for the instance by ID, I've deleted them. But even after using ipcrm to remove the now-dead instance's shared memory segments, they are still being reported as being referenced by the O/S (ipcs status 'D'), preventing the informixserver from being restarted. All the oninit processes for the instance were killed. At this point is there an alternative to using the Microsoft solution and rebooting the host?

red_valsen wrote: > On Mar 6, 9:34 am, theBP <th...@Usenet-News.Net> wrote: >> red_val...@yahoo.com wrote: >>> I have 3 informixservers running on a single host. One of the >>> instances has experienced an assertion failure and is now >>> unresponsive, so I'd like to kill its oninit processes and attempt to >>> restart it. But running onstat -g ses only reports a PID of 0 for the >>> oninit sessions. How do I determine which oninit processes are >>> associated with which instances? >>> Running IDS 9.40FC8 on HPUX11. >> onstat -g glo>> >> Establish the main CPU VP. >> >> Get an onstat -g stk all first, and an onstat -a for diagnostics. > > Thank you. > > Now that I've been able to identify the oninit processes for the > instance by ID, I've deleted them. But even after using ipcrm to > remove the now-dead instance's shared memory segments, they are still > being reported as being referenced by the O/S (ipcs status 'D'), > preventing the informixserver from being restarted. All the oninit > processes for the instance were killed. At this point is there an > alternative to using the Microsoft solution and rebooting the host? Weird... Can you check if there is any "onstat" hanging around that doesn't belong to any of the other instances? -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently...

On 6 Mar, 18:56, red_valsen <red_val...@yahoo.com> wrote: > On Mar 6, 9:34 am, theBP <th...@Usenet-News.Net> wrote: > > > red_val...@yahoo.com wrote: > > > I have 3 informixservers running on a single host.  One of the > > > instances has experienced an assertion failure and is now > > > unresponsive, so I'd like to kill its oninit processes and attempt to > > > restart it.  But running onstat -g ses only reports a PID of 0 for the > > > oninit sessions.  How do I determine which oninit processes are > > > associated with which instances? > > > > Running IDS 9.40FC8 on HPUX11. > > > onstat -g glo> > > Establish the main CPU VP. > > > Get an onstat -g stk all first, and an onstat -a for diagnostics. > > Thank you. > > Now that I've been able to identify the oninit processes for the > instance by ID, I've deleted them.  But even after using ipcrm to > remove the now-dead instance's shared memory segments, they are still > being reported as being referenced by the O/S (ipcs status 'D'), > preventing the informixserver from being restarted.  All the oninit > processes for the instance were killed.  At this point is there an > alternative to using the Microsoft solution and rebooting the host? There should be an option to ipcs to determine what is attached to the segment. (maybe -o or -p)