svn-1281 feedback…

[Dave.B]

I upgraded from the previous svn on Friday morning. Everything appeared to work well then. I’ve just got back from two days away (NSLU left running).

My Soundbridge wouldn’t connect to Firelfy, and although I could access the Firefly web interface, it didn’t display any information about songs etc… and clicking the Run Scan buttons did nothing.
So I restarted the NSLU2. It came back up OK and once Firefly started I could get the Soundbridge to connect and search/browse perfectly… but this only lasts for around 1-2 minutes before the whole lot becomes totally unresponsive again. I’ve done this 3 times now, with the same result each time.
The only other thing I’ve tried is to start Firelfy using the -r switch (using the instructions Ron emailed to me a few weeks ago), but this just says “Error reading /opt/etc/mt-daapd.conf”, even though I supplied the correct path “/opt/etc/mt-daapd/mt-daapd.conf”.

I’m fresh out of ideas, and I wish I had’t upgraded now as this is a showstopper. Can I downgrade to the previous version?

Bear in mind that if you want diagnostic logs or anything, you’ll need to tell me how (in words of one syllable or less!)
Thanks
Dave

[Dave.B]

OK, I might have been a bit hasty on this one. I powered the slug down using the power button, shut down the disk, then powered everything back up again.

It’s now run for over 14 minutes (much longer than the last 4 times I started it) and is just about to (hopefully) serve up it’s third song………… yes, it did.

So there was a problem, but a minor one it seems.
Forget I said anything.

[cromei]

I’ve had similar problems with my NSLU as well where the entire Slug will lockup. I can not get it to reproduce tho’. Everytime I have the log on it works fine….
Best reproduction method has been add songs to the scan drives and start a full scan. This has locked it up about 20% of the time. I will continue to try and get a log to send in.
Craig

[Dave.B]

OK, it does look like something might be wrong. It was fine for 53 minutes and now it has become unresponsive again.
The file scanner is apparently active, although the disk isn’t doing anything, but the little rotating arrow icon top left of the screen is rotating. The Soundbridge will not connect either, as expected.

The slug itself appears fine as I can read/write to the disk through Windows Explorer with no trouble.

Any suggestions?

[rpedde]

@Dave.B wrote:

OK, it does look like something might be wrong. It was fine for 53 minutes and now it has become unresponsive again.
The file scanner is apparently active, although the disk isn’t doing anything, but the little rotating arrow icon top left of the screen is rotating. The Soundbridge will not connect either, as expected.

The slug itself appears fine as I can read/write to the disk through Windows Explorer with no trouble.

Any suggestions?

This sounds like a deadlock, but it’s one I haven’t seen yet. I’ll try and track it down.

If you have the disk space to run -d9 and get a log of when it hangs, that would be a big help.

— Ron

[Dave.B]

Ron, I’m only too willing to generate a log, but as you know, I’m struggling like a complete doofus with all this!

My guess at the moment is:
specify a path for the logfile on the config page
Start the server using

/opt/sbin/mt-daapd -d9 /opt/etc/mt-daapd/mt-daapd.conf

If so, see my problem above about it seemingly not seeing the conf file.

If you have the disk space to run -d9…

I have 73% of 300GB free disk space – I assume the log will nowhere near that big?!

[Dave.B]

OK, I seem to have really ballsed it up now. I thought I’d try and downgrade to 1249, which was working fine for me last week.
Just doing the ipkg install mt-daapd…… etc wouldn’t allow me to downgrade, so I used the instructions here:
http://forums.fireflymediaserver.org/viewtopic.php?t=982&highlight=uninstall
to uninstall 1281, which reported that it worked, but didn’t seem to do much otherwise (no disk activity).
It then allowed me to install 1249, but again, very little disk activity. When prompted to start the daemon, I typed /opt/etc/init.d/S60mt-daapd. I get the prompt back again, but nothing happens, no disk activity, and no server…

So I lay myself at your mercy… please help me!

[Dave.B]

Note the times between these two posts… that’s how long it took me to figure out how to generate a log 😳 At least I can do it now…

2006-07-02 22:09:51: Starting with debuglevel 9
2006-07-02 22:09:51: Starting rendezvous daemon
2006-07-02 22:09:51: mDNSPlatformTimeNow went backwards by 6 ticks; setting correction factor to 6
2006-07-02 22:09:51: get_ifi_info
2006-07-02 22:09:51: intf name=lo AF=2, flags=00000002
2006-07-02 22:09:51: intf name=ixp0 AF=2, flags=00000002
2006-07-02 22:09:51: SetupInterfaceList
2006-07-02 22:09:51: mDNS_RegisterInterface: InterfaceID 0005CC80 192.168.2.77 not represented in list; marking active and retriggering queries
2006-07-02 22:09:51: Adding 0005CC90 NSLU2.local. (Addr) to active record list
2006-07-02 22:09:51: Adding 0005CF3C 77.2.168.192.in-addr.arpa. (PTR) to active record list
2006-07-02 22:09:51: Not creating HINFO record: platform support layer provided no information
2006-07-02 22:09:51: SetupOneInterface: ixp0 192.168.2.77 Registered
2006-07-02 22:09:51: select(7, 0.000976)
2006-07-02 22:09:51: mDNS_Execute
2006-07-02 22:09:51: select(7, 0.223632)
2006-07-02 22:09:51: Starting signal handler
2006-07-02 22:09:51: Signal handler started
2006-07-02 22:09:51: Opening database
2006-07-02 22:09:51: Executing: select value from config where term=’version’
2006-07-02 22:09:51: Old database version.
2006-07-02 22:09:51: Results: 12
2006-07-02 22:09:51: Error opening db: Wrong db schema. Use mtd-update to upgrade the db.
2006-07-02 22:09:51: Processing rendezvous message
2006-07-02 22:09:51: Stopping mDNS
2006-07-02 22:09:51: Exiting
2006-07-02 22:09:51: mDNS_Close: RR Cache was using 0 records, 0 active
2006-07-02 22:09:51: mDNS_Deregister_internal: Deleting record for NSLU2.local. (Addr)
2006-07-02 22:09:51: mDNS_Deregister_internal: Deleting record for 77.2.168.192.in-addr.arpa. (PTR)
2006-07-02 22:09:51: mDNS_Close: No deregistering records remain
2006-07-02 22:09:51: mDNS_Close: mDNSPlatformClose
2006-07-02 22:09:51: mDNSPlatformTimeNow went backwards by 5 ticks; setting correction factor to 11
2006-07-02 22:09:51: mDNS_DeregisterInterface: Last representative of InterfaceID 0005CC80 deregistered; marking questions etc. dormant
2006-07-02 22:09:51: mDNS_Close: done
2006-07-02 22:09:51: Finished with status 0, result 0
2006-07-02 22:09:51: Stopping signal handler

OK, the problem seems obvious now, but how exactly do I “Use mtd-update to upgrade the db”? Can I just delete the database files? I have 3:
songs.db
songs.db-7aq0xqivhujbowij84tv
songs.db-7aq0xqivhujbowij84tv-journal

Thanks in advance
Dave

[grommet]

svn-1281 (Win32) died on me again:

2006-07-02 : Finished streaming file to remote: 5489441 bytes
2006-07-02 : Finished streaming file to remote: 7198113 bytes
2006-07-02 : Out of pseudo file handles. See ya
2006-07-02 : Aborting

So, it’s crashing more than hanging… Is that good? 😉

[rpedde]

@grommet wrote:

svn-1281 (Win32) died on me again:

2006-07-02 : Finished streaming file to remote: 5489441 bytes
2006-07-02 : Finished streaming file to remote: 7198113 bytes
2006-07-02 : Out of pseudo file handles. See ya
2006-07-02 : Aborting

So, it’s crashing more than hanging… Is that good? 😉

The phandles thing is a win32 issue… a socket isn’t being closed. I just need to find out where.

Dave.B:

yes, just delete the database files and it will regenerate them. Then just run it with a debug log until it dies, and email me the zipped log to ron at pedde.com.

thanks.

— Ron

[Author]

Posts