Error message during system test, I think expected, but...

[Myrna van Lunteren <m....@gmail.com>]

Hi,

During the run against 10.8.3.0 of the nstest client/server system
test, which retries to connect to a backed-up database with a bad
password, generating XBM06 errors, I also see the following in
derby.log:

Mon Nov 05 03:21:36 PST 2012
Thread[derby.index-stat-thread,5,derby.daemons] {istat} shutting down
daemon, 50 consecutive errors seen
java.lang.IllegalStateException: degraded state
        at org.apache.derby.impl.services.daemon.IndexStatisticsDaemonImpl.processingLoop(Unknown
Source)
        at org.apache.derby.impl.services.daemon.IndexStatisticsDaemonImpl.run(Unknown
Source)
        at java.lang.Thread.run(Thread.java:736)

Note that while there are indeed repeated XBM06 errors regarding
connect to restored database, the test merrily continues to prattle to
the active database without any errors.
On checking the saved output from my 10.8.2.2 and 10.9.1.0 run, I can
see the same behavior there.

I think this is likely expected behavior, but can anyone confirm/correct?

Thx,
Myrna

Re: Error message during system test, I think expected, but...

[Kristian Waagan <kr...@apache.org>]

On 13.11.2012 20:28, Myrna van Lunteren wrote:
> Hi,
>
> During the run against 10.8.3.0 of the nstest client/server system
> test, which retries to connect to a backed-up database with a bad
> password, generating XBM06 errors, I also see the following in
> derby.log:
>
> Mon Nov 05 03:21:36 PST 2012
> Thread[derby.index-stat-thread,5,derby.daemons] {istat} shutting down
> daemon, 50 consecutive errors seen
> java.lang.IllegalStateException: degraded state
>          at org.apache.derby.impl.services.daemon.IndexStatisticsDaemonImpl.processingLoop(Unknown
> Source)
>          at org.apache.derby.impl.services.daemon.IndexStatisticsDaemonImpl.run(Unknown
> Source)
>          at java.lang.Thread.run(Thread.java:736)
>
> Note that while there are indeed repeated XBM06 errors regarding
> connect to restored database, the test merrily continues to prattle to
> the active database without any errors.
> On checking the saved output from my 10.8.2.2 and 10.9.1.0 run, I can
> see the same behavior there.
>
> I think this is likely expected behavior, but can anyone confirm/correct?

Hi Myrna,

This is the first time I see this error reported. Do you have a log 
telling what the errors encountered by the istat daemon are?
You may have to enable istat logging and/or tracing to see that.

While the istat daemon will be shut down for the database, this will not 
affect other activities in the database. However, I would expect that if 
the istat daemon fails on 50 consecutive attempts, there is a 
non-trivial problem in the system, or a bug in the daemon itself.

This feature was added to stop the daemon from eating up the system 
resources in cases where it keeps failing for some reason.


Regards,
-- 
Kristian

>
> Thx,
> Myrna