Recording service locks up

[hondophred]

Hi,

The recording service stops processing new recordings and the application becomes incredibly slow to respond to the remote (1 -2 minutes between a button press and it working.)and commercial skipping ceases

I'm including the logs. This problem is intermittent in other words the recording usually works fine for a while 2-3 days maybe and then locks up usually when i try to record 2-3 programs at once. but not always. In other words i can record 2-3 programs at once.

noticed the line...

Failure in ChannelFactory::loadProgrammeByOID(): The database file is locked
database is locked

--- in the recording service logs but could not find it in a search of the forums.

If I restart the recording service thigs are happy again.

[sub]

Basically something caused your database to lock up, but there isnt enough info in the logs to tell why.

To help me identify the cause, you'll need to set the logging level back to 'debug' and also set <LogSQL>true</LogSQL> in config.xml. They next time it happens, zip and attach a complete set of logs.

Are you running any plugins?

[hondophred]

Will do...

I'm not running any plugins beyond what comes with the standard install.

[hondophred]

Some new logs

[sub]

Do you use the GB-PVR front end on your PC as well as your MVP?

[hondophred]

Yes i use the GBPVR front end on the server. It is the primary way i use the App.

On occasion i use MVP and GBPVR in client mode on another PC but they haven't been in use when the recording service locks up.

specs on my machine...
amd FX60 processor
1024 mb of ram
nforce 4 motherboard
3 hauppauge pvr150 boards (pci)
a dts connect soundcard (pci)
ATI x1900 xt video card

avg for anti-virus

little else as far as software is installed. the XP install is about 3 weeks old.

[sub]

Can you disable the MVP server, then reproduce the problem and repost a set of logs? I think the useful bits of info are in the GBPVR.exe logs, but when the MVP server is running, this only contains logging from the MVP server.

[hondophred]

OK so last night I tried replicating the error w/o an MVP service running. and wouldn't you know it my intermittent error has now become shy.

I'll keep on the lookout for the behaviour. If or when it happens again I'll report back with some fresh logs.

Of course now this has me suspecting that it was the MVP service causing the db locking issue. I was considering setting the read_uncommitted flag to true in the SQLite. Any known hazards on your end (other than the usual dirty read concerns)?

[sub]

Of course now this has me suspecting that it was the MVP service causing the db locking issue.

I couldnt see anything in that MVP logs, but you never know...

The next release produces separate MVP and PC logs.

I was considering setting the read_uncommitted flag to true in the SQLite. Any known hazards on your end (other than the usual dirty read concerns)?

I've not tested this way, so cant really comment.