FireFly Media Server › Firefly Media Server Forums › Firefly Media Server › Nightlies Feedback › 1696 caused 100% CPU usage on XP SP2
- This topic has 21 replies, 8 voices, and was last updated 17 years ago by indierocksteve.
-
AuthorPosts
-
07/11/2007 at 3:42 PM #14200dantidoteParticipant
I have zero files that must be transcoded, but it’ll still hover at 100%. Even when no one is connected. Weird.
09/11/2007 at 3:04 AM #14201rpeddeParticipant@dantidote wrote:
I have zero files that must be transcoded, but it’ll still hover at 100%. Even when no one is connected. Weird.
Just as a data point, do either of you see high cpu if you disable background scanning by setting the scan_interval to 0?
11/11/2007 at 8:43 AM #14202AnonymousInactiveI turned off scanning completely and it’s still sitting there using 100%- the last activity was an hour ago.
11/11/2007 at 10:59 PM #14203rpeddeParticipant@redbeard wrote:
I turned off scanning completely and it’s still sitting there using 100%- the last activity was an hour ago.
Okay, good. I appreciate the help. I’m going to finish up working on playlists, and once I’ve finished that, I’ll try and nail this down.
— Ron
14/11/2007 at 3:32 PM #14204AnonymousInactiveI am having this pegged CPU utilization issue as well.
I am running:
2007-11-14 10:19:18 (3c77e18b): Firefly Version svn-1696: Starting with debuglevel 2
2007-11-14 10:19:19 (3c77e18b): Plugin loaded: w32-event/svn-1696
2007-11-14 10:19:19 (3c77e18b): Plugin loaded: ssc-wma/svn-1696
2007-11-14 10:19:19 (3c77e18b): Plugin loaded: ssc-ffmpeg/svn-1696
2007-11-14 10:19:19 (3c77e18b): Plugin loaded: rsp/svn-1696
2007-11-14 10:19:19 (3c77e18b): Plugin loaded: daap/svn-1696
2007-11-14 10:19:19 (3c77e18b): Starting rendezvous daemon
2007-11-14 10:19:19 (3c77e18b): Building drive mapping table from C:Program FilesFirefly Media Servermapping.ini
2007-11-14 10:19:19 (3c77e18b): Initializing database
2007-11-14 10:19:28 (3c77e18b): Starting web server from C:Program FilesFirefly Media Serveradmin-root on port 9999
2007-11-14 10:19:28 (3c77e18b): Registering rendezvous names
2007-11-14 10:19:28 (3c77e18b): Serving 12514 songs. Startup complete in 10 seconds.Running on Windoze XP Pro with SP2, AMD 1.4 GHZ, and 1.2GB RAM.
I have scanning disabled and no one is accessing the server as this is on my home private network.
I start the server up with no clients hitting it, come back in a few hours and the server is at 100%. Memory utilization is only at about 8,000k on firefly.exe and 5,168k on FireflyShell.exe
Update: My server was running for 23 hours at 1-2% CPU utilization. On the 23rd hour, firefly.exe flatlined at 95% utilization. I have logging set to the highest level and no new events are recorded from the log events seen above.
About two hours ago a connected two clients to the server, an itunes client and a Banshee client. Both clients connected to the server, loaded up the playlist and then I closed down the clients.
17/11/2007 at 3:11 AM #14205AnonymousInactiveI don’t know if this will help at all or if you’ve already got an idea of what’s going on but:
I was just sitting around today and had noticed that one of the cores on my serving computer as doing the usual 100% use thing. A few minutes later I suddenly noticed it stopped and this was the error that occurred when it stopped:2007-11-16 19:06:50 (7d09ff67): Thread 20: could not read: unknown internal error
The thread it is in reference to is:
2007-11-16 17:36:00 (7d09ff67): Request: GET daap://146.244.184.248:9999/server-info HTTP/1.1
2007-11-16 17:36:01 (7d09ff67): Request: GET /content-codes HTTP/1.1
2007-11-16 17:36:01 (7d09ff67): Request: GET /databases HTTP/1.1
2007-11-16 17:36:02 (7d09ff67): Request: GET /databases/1/items?type=music&meta=dmap.itemid,dmap.itemname, HTTP/1.1
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songalbumartist
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songcategory
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songextradata
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songcontentdescription
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songlongcontentdescription
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songkeywords
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.is-podcast
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.mediakind
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.series-name
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.network-name
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.episode-num-str
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.episode-sort
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.season-num
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: daap.songgapless
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.gapless-enc-del
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.gapless-heur
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.gapless-enc-dr
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.gapless-dur
2007-11-16 17:36:02 (7d09ff67): Unknown meta code: com.apple.itunes.gapless-resy
2007-11-16 17:36:03 (7d09ff67): Request: GET /databases/1/containers?meta=dmap.itemid,dmap.itemname, HTTP/1.1
2007-11-16 17:36:03 (7d09ff67): Unknown meta code: com.apple.itunes.is-podcast-playlist
2007-11-16 17:36:03 (7d09ff67): Unknown meta code: com.apple.itunes.special-playlist
2007-11-16 17:36:04 (7d09ff67): Request: GET /databases/1/containers/1/items?type=music&meta=dmap.itemkind,dmap.itemid,dmap.containeritemid HTTP/1.1So a good hour and a half difference before the thread ending happened. I’m fairly certain that’s the exact error at the exact time it stopped, as when my cpu usage is at 100% the fan on my server kicks up, and it suddenly kicked down shortly after that error ending the thread.
edit: just realized this was horribly table breaking
18/11/2007 at 9:53 AM #14206mbakerParticipantHiya. I’m having the same problem with this build but seeing some different symptoms in the log file.
I’ve got debug set to 2 and am getting lots and lots of the following:
...
2007-11-18 08:35:32 (d374dfdb): Updating playlists
2007-11-18 08:35:33 (d374dfdb): Updating playlists
2007-11-18 08:35:33 (d374dfdb): Updating playlists
...When I connect to it using my Roku, it’ll usually work briefly before crashing with the following error:
2007-11-18 08:37:09 (7808c3cf): Error: enum_begin failed (error 1): Misc SQL Error: database is locked
So basically, it’s dead Ron…
Good luck with the DB rewrite! Interested to see that with this version you’ve upgraded to SQL3 (I’ve been running on an extremely stable 14xx version for months).
19/11/2007 at 3:46 AM #14207rpeddeParticipant@mbaker wrote:
Hiya. I’m having the same problem with this build but seeing some different symptoms in the log file.
I’ve got debug set to 2 and am getting lots and lots of the following:
...
2007-11-18 08:35:32 (d374dfdb): Updating playlists
2007-11-18 08:35:33 (d374dfdb): Updating playlists
2007-11-18 08:35:33 (d374dfdb): Updating playlists
...When I connect to it using my Roku, it’ll usually work briefly before crashing with the following error:
2007-11-18 08:37:09 (7808c3cf): Error: enum_begin failed (error 1): Misc SQL Error: database is locked
So basically, it’s dead Ron…
Good luck with the DB rewrite! Interested to see that with this version you’ve upgraded to SQL3 (I’ve been running on an extremely stable 14xx version for months).
It’s supported sqlite2 and sqlite3 for a long time now. sqlite3 was kind of changing a moving target though, as they kept changing the threading rules.
But mostly it’s stalbleish now.
Is this a windows box? Because that looks more like virus scanner than anything else.
– Ron
19/11/2007 at 11:19 AM #14208mbakerParticipantIt is a Windoze box, but it’s a server (Win Server 2003) without a virus scanner running. Haven’t had a problem like this up to now with any of the older nightlies (but then I haven’t upgraded in a long while).
Have just moved the iTunes files from the local machine to a NAS (Terastation Live where I’m hoping to migrate Firefly to at some point soon so I can retire the server) but otherwise that’s the only change.
I remember you commenting on the Sqllite 3 “moving target” previously. Glad to hear it’s settling. That I didn’t know shows how stable my 14xx has been!
Did some more testing last night/today. Turned off scanning last night and then started Firefly and left it overnight without connecting. It burned hard on the startup scan generating loads of Updating Playlist messages. This morning I connected and it played 2 songs before dying like this.
...
2007-11-18 19:35:56 (d374dfdb): Updating playlists
2007-11-18 19:35:57 (d374dfdb): Updating playlists
2007-11-18 19:36:02 (d374dfdb): Updating playlists
2007-11-18 19:36:12 (d374dfdb): Scanned 14959 songs (was 14959) in 452 seconds
2007-11-19 10:25:00 (e015e054): Session 0: Streaming file '01 Silver Lining.m4a' to 192.168.0.6 (offset 0)
2007-11-19 10:28:21 (e016f916): Session 0: Streaming file '02 Close Call.m4a' to 192.168.0.6 (offset 0)
2007-11-19 10:29:10 (d374dfdb): Rescanning database
2007-11-19 10:29:11 (d374dfdb): Starting playlist scan
2007-11-19 10:33:16 (5e28b214): Error: enum_begin failed (error 1): Misc SQL Error: database is lockedIt looks like it’s bumping into itself when trying to get a new song while scanning’s taking place. Seems pretty reproducible. Would a level 9 dump help?
Cheers,
Mark21/11/2007 at 4:39 AM #14209rpeddeParticipant@mbaker wrote:
It is a Windoze box, but it’s a server (Win Server 2003) without a virus scanner running. Haven’t had a problem like this up to now with any of the older nightlies (but then I haven’t upgraded in a long while).
Have just moved the iTunes files from the local machine to a NAS (Terastation Live where I’m hoping to migrate Firefly to at some point soon so I can retire the server) but otherwise that’s the only change.
I remember you commenting on the Sqllite 3 “moving target” previously. Glad to hear it’s settling. That I didn’t know shows how stable my 14xx has been!
Did some more testing last night/today. Turned off scanning last night and then started Firefly and left it overnight without connecting. It burned hard on the startup scan generating loads of Updating Playlist messages. This morning I connected and it played 2 songs before dying like this.
...
2007-11-18 19:35:56 (d374dfdb): Updating playlists
2007-11-18 19:35:57 (d374dfdb): Updating playlists
2007-11-18 19:36:02 (d374dfdb): Updating playlists
2007-11-18 19:36:12 (d374dfdb): Scanned 14959 songs (was 14959) in 452 seconds
2007-11-19 10:25:00 (e015e054): Session 0: Streaming file '01 Silver Lining.m4a' to 192.168.0.6 (offset 0)
2007-11-19 10:28:21 (e016f916): Session 0: Streaming file '02 Close Call.m4a' to 192.168.0.6 (offset 0)
2007-11-19 10:29:10 (d374dfdb): Rescanning database
2007-11-19 10:29:11 (d374dfdb): Starting playlist scan
2007-11-19 10:33:16 (5e28b214): Error: enum_begin failed (error 1): Misc SQL Error: database is lockedIt looks like it’s bumping into itself when trying to get a new song while scanning’s taking place. Seems pretty reproducible. Would a level 9 dump help?
Cheers,
MarkActually, not really. I spend a whole day a couple weekends ago trying to find it and didn’t see it. I’m in the middle of completely redoing the db backend anyway right now, so by the time I’m done, the db stuff will be radically simplified and the problem will probably go away anyway. So I think it’s going to be a non-issue next nightlies anyway.
Thanks for the offer of help, though.
— Ron
-
AuthorPosts
- The forum ‘Nightlies Feedback’ is closed to new topics and replies.