Anyone see this? I was trying to troubleshoot appserver not starting for an hour today, only to find out that it was actually being started correctly.
Connecting to Progress AdminServer using rmi://localhost:20931/Chimera (8280)
Searching for as12345 (8288)
Connecting to as12345 (8276)
Unable to start as12345 (8297)
It was being started correctly, but Progress thought it would be fun to just not say so.
See: http://knowledgebase.progress.com/articles/Article/000054033
How aggravating. You'd think the error message "unable to start" would say "We really don't know if your app server is running or not. It may or may not be running but it is not responding to a ping.". Just give it to me straight, Progress...
We have OE 11.6 and apparently this issue has been known for while, maybe classic appserver bugs are less of a priority than the shiny new PAS stuff?
Progress moved this article.
The problem may be specific to Windows x64. I opened a ticket with Progress to try to sort this out. We never saw it on Windows x86 or HP-UX.
Would be nice to hear if anyone else is getting bogus failure messages when starting appservers. It seems inadvisable but perhaps this is something that most OE admins know that they can just ignore.
Actually there is another KB article on the problem too:
knowledgebase.progress.com/.../000051987
On further investigation, it appears that this has been a problem since 11.1. (see www.progresstalk.com/.../).
It is unfortunate that Progress lets us spend our time chasing down an error message that isn't an error at all.
In our case we are migrating away from HP-UX where this error message was *never* seen on any of our five servers, to Windows x64 where it happens *every* time we use asbman to start a broker. I don't think I buy the "timing issue" argument or the problem would be intermittent instead of consistent.
I sorted this out with Progress tech support. Apparently there should not normally be a message stating either way whether the appserver started successfully or failed to start. The error you should be looking for is simply:
Starting as12345. Check status. (8296)
However, in our case we were seeing the "unable to start" message when attempting to start the appserver broker.
Per tech support, the first thing to try is starting the default app server that is defined in ubroker.properties. It is called asbroker1. We issued the start command and everything worked as expected:
Starting asbroker1. Check status.
Then we eliminated the differences between the configuration of that broker and the one that says "unable to start" (a bogus message given that it actually starts).
By the process of elimination (removing things like logging preferences, autostart preferences, etc) we finally commented out the setting of the PROPATH and that finally gave us the right startup message (vs "unable to start").
So for anyone else getting the bogus "unable to start" message, you should first check the behavior when starting asbroker1 and then try to isolate the difference between your broker and asbroker1. Hope this helps.