Knut Forkalsrud escribió:
> Daniel López wrote:
>> Knut, you seem to have a setup quite similar to ours, are you using 3.1
>> already or staying if 3.0? If you are using 3.1, are you experiencing
>> similiar issues or is it just me? ;).
> We're on 3.0, but one of my short term objectives is to get us migrated
> to 3.1. I think we're almost there, although when I saw your mails
> about watchdog problems I paused for an extra check :-). The problems
> you describe haven't struck us, at least not so far. Then we haven't
> had the JVM freeze on us either. We probably will not run into it
> either as long as we insist on different OS users/watchdogs per JVM.
> The confusion over watchdog port caught us by surprise as well, it took
> a second or third close reading of the documentation to find that one.
Well, our problem is that we seem to have some flaky
databases/connectivity, and Oracle servers seem to be quite creative
regarding their JDBC connections and how they are left after a DB shuts
down. So it's usually not a JVM freeze but the datasources behaving
wildly. I usually try to restart the context to see if that fixes it,
but it usually involves restarting the whole JVM. That's also one of the
reasons we use various instances grouped by database.
Different watchdogs is also not an option for us, as we are quite
constrained in the resources side, and doubling the number of JVMs would
be too much, even if the watchdog process does not use much memory. The
JVM itself already has an overhead that we cannot assume per-instance. I
had to fight for one whole month so the admin guys would give me 1Gb RAM
extra instead of 1Gb virtual memory... Physical RAM should work the
same as virtual memory paged on disk, right? oh, yeah, specially for a
Garbage collector JVM... sigh :).
I know it's not an ideal situation but that's the conditions we have to
Regarding your explanation, Scott, that makes sense as I usually try to
stop the instance normally, and just after that, if the process is still
alive I kill it "manually". It might happen that if the JVM is in the
process of stopping itself but it takes too long, I kill it in the
middle of the stop action... would that help to explain the watchdog
stop process to fail and the inconsistent state?
resin-interest mailing list