Cancel during recording - db locked, file still there.

[martint123]

I was recording the F1 qualifying today and was watching it directly on the TV.
When it ended I used the PCH to cancel the current recording as I'd added loads of padding. I knew something was wrong when the delete popup stayed stuck on the screen for a minute - then it went back to the pending menu and it was still showing as recording. I cancelled again and it appeared normal. I then went to the new screen and deleted the recording.

I noticed when doing a backup that the recording it was still on disk. Looking at "new" recordings, it seems to have been deleted OK from teh database.


Started recording around 12:13 (inc padding)
Cancel given around 14:02:19 and second at 14:03:08

To my untrained eye, it looks like it started recording again at
14:02:54:984 there was only one file on disk and it was the big one from the beginning.

error at 14:02:52 with database locked that probably caused it all, but why does it happen? and so infrequently?

cheers
Martin

[pBS]

same here, but it happens pretty constantly here...
http://forums.nextpvr.com/showthread.php?t=36764

what epg are you using? was thinking maybe something in the db..
i tried everything and can't figure out where it's coming from..
[regular database locked errors,but clears up after about 30 secs]
only happens on in-progress recordings...

[martint123]

Yes, very similar.
My EPG comes from the Radio Times, via XMLTVGUI and the update time is nowhere near the recording time. No plugins. Blue Skin. Most of the use of gbpvr is via the PCH and that is the only place I've seen the problem - BUT that is probably where 99% of the action comes from. System is set up for going to S3 standby when not in use. Gets rebooted about every three weeks.
Is there an easy way to regularly monitor if the db is locked - something running separately and maybe logging status??
I've had a sweep up of my recording directory and found around 35gb of recordings that have been deleted from the database, but still existed on the drive (probably from the above cause).

[pBS]

hmmm,not much in common at all...'cept no plugins and blue skin...
i never send it to sleep, primarily use it for watching, epg is from mc2xml with ultraxml, only leftover files are the ones from restarting recording, [which i usually remove promptly for space]

hmm, how big is your db? how long since last fresh install?
what remote? MCE2k5 here...on rebuild i'm goin with pro tho.. do you do updates regularly?
it only seems to happen for me for a few seconds at a time... usually self clears...
checked and vacuumed db...
no problems in epg data for long time...


just trying to find a commonality..

[martint123]

Even further apart I guess.

db is around 15mb
Last fresh install of gbpvr and win XP PRO would be around a year ago, but database was copied back down, so not fresh.
No remote, server is 99% server only, not used for viewing.
XP Pro SP2, no updates since SP2.
Vacuum maybe once in a while, certainly longer than a year ago.
No problems with EPG for a long time as well.
I restart the recording service each time the system wakes from standby.

For me the db locked problem is not common, but very noticable when it happens. I can find no cause or solution. Since the last one I posted at the top of this thread, I have no rebooted and it hasn't done it since.

So very little commonality in fact :confused:

[pBS]

do you have autoremovemissingrecordings set to false?
my problems were caused by missing recordings bogging down the refresh of recordings list, keeping the db busy...
tho i have a feeling slow to respond servers or slow to spin up usb drives may cause similar, tho probably more intermittent, problems...
my missing recs were on a server that didn't exist anymore so it may have a different timeout.
mine's fixed completely now tho..

[martint123]

do you have autoremovemissingrecordings set to false?

I have mine set to false. Mainly because I have a 750gb Samsung drive that once in a blue moon will not spin up from hibernate. This results in my DB getting wiped. So what I do every so often is set it to "true" and fire the system up and let it tidy itself up and then set it back to false.

I used to have powersaving set to let the drives spin down after 30 minutes, but I've turned that off now and DB locking seems much reduced. (but don't even think of asking why!).