Bug 8879 - update_begin: bad state shows during upgrades
: update_begin: bad state shows during upgrades
Status: CLOSED FIXED
Product: SB Boom
Classification: Unclassified
Component: Hardware
: 33
: PC Windows XP
: -- major (vote)
: 7.4.0
Assigned To: Felix Mueller
http://forums.slimdevices.com/showthr...
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2008-07-25 20:35 UTC by Caleb Crome
Modified: 2009-10-05 14:30 UTC (History)
3 users (show)

See Also:
Category: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Caleb Crome 2008-07-25 20:35:02 UTC
I've seen this too, but I assumed it had to do with my debug builds.  It seems to happen more often when I'm at the edge of wifi reception, though I have not reproduced it reliably.


On Fri, Jul 25, 2008 at 2:59 PM, max.spicer <max.spicer.3d4jhz1217023202@no-mx.forums.slimdevices.com> wrote:


    I just updated to the latest on the boom branch and saw some odd
    behaviour when I restarted the server.  Both my boom and my transporter
    went through the fw upgrade process at least three times, failing each
    time.  On my boom, I saw the error 'update_begin: bad state' in the
    middle of the progress bar stage.  I saw the transporter flash up an
    error too, but sadly wasn't close enough to the display to read it.
    After the third (ish) attempt, the updates completed successfully.  At
    the same time as this, my wireless network appeared to go screwy - my
    open ssh session dropped and I couldn't contact external web pages.

    Now obviously, the fw failures could have been simply due to my
    wireless network going odd.  But it's normally very stable.  In fact, I
    can't remember it behaving like that before.  Could the wireless oddness
    have been caused by the fw updates doing nasty packet storm things or
    something?

    Thought it was worth posting here in case there is a problem with the
    update.

    Max


    --
    max.spicer
Comment 1 Chris Owens 2008-07-28 09:09:38 UTC
QA to try upgrading multiple players, on congested networks, and at limits of range.

Customer comments would be welcome as well.
Comment 2 James Richardson 2008-08-05 21:47:54 UTC
I have been unable to reproduce this either at MV or on my home / other networks I have access to.  Bumping this to 7.3
Comment 3 Max Spicer 2008-08-13 11:35:38 UTC
I'm still seeing this occasionally - it happened just now, for example.  I didn't notice if an error was displayed on the screens, but my wireless network went offline for a short period when I restarted squeezecenter having just updated to the latest svn that included fw upgrades.
Comment 4 Max Spicer 2008-08-15 02:01:52 UTC
This happened again with all the fw updates included with Boom v30/31.  This time I was deliberately watching the Boom.  The fw update stage began and the bar got to approx 1/3 full.  It then hung for a while and afterwards the screen went blank.  My wireless network hung at this point, causing my ssh session to my server to be dropped.  Boom displayed the update_begin: bad state message.  I think at this point my Transporter restarted as I heard it click (audio outs turning on/off, probably).  The transporter could then be turned on and was connected to sc.  My boom was at the fw's select source menu and I had to make it connect to sc manually.  It then did another fw update, then displayed Software Update.  This completed and it did another fw update.

My wireless network isn't the fastest, but it's normally stable.  I've currently got two Booms (pqp3 and a pqp1), a Transporter, a SB2, a SBR and a SBC powered on and connected to SC.  In the past I've probably had other devices connected as well when I've seen this bug, but they are currently physically powered off.

It could well be that this happens every time I get a fw update as I often update and restart remotely or simply don't pay that much attention.  Worth reconsidering for 7.2?
Comment 5 Max Spicer 2008-08-20 15:01:18 UTC
I've just done the latest fw upgrade - transporter 62 and boom 32 etc.  The same happened as before, but this time I noticed that the transporter failed first and displayed an error message ending with 'slimproto'.  I would bet it was probably 'nb alloc failed in slimproto', which has been reported on the forums.  The transporter rebooted, at which point the boom displayed the 'update begin: bad state' message.  As before, my wireless network hung during this process.  It's possible that this only happens when the transporter is in the equation.
Comment 6 Felix Mueller 2008-10-20 03:07:19 UTC
From what I read here I start to believe it has to do with multiple players requesting firmware upgrades at the same time, which is something that probably did not happen that often before SBR / Boom and before we had SBR / Boom set to do auto-upgrade without manual interaction.

A player shows this error message if there is some packet arriving in the wrong order. Could it be that SC / SN gets confused about what player is in what upgrade stage if multiple players are upgrading at the same time?

Did QA already try this with let's say multiple Booms connected to one SC?

BTW: I am not sure it is a major bug as it actually prevents player firmware from getting corrupt. Maybe we just need a better wording for the error message?

Sean: What do you think? You are more familiar with the upgrade process. Thanks.
Comment 7 Max Spicer 2008-10-20 03:52:01 UTC
Maybe not major, but it's pretty annoying every time it takes out my wireless network.  I have all my players set to auto-upgrade fw.
Comment 8 Felix Mueller 2008-10-20 04:27:24 UTC
Ok, I didn't realize the firmware upgrade process is actually killing your wireless network which seems really strange. I thought because of your wireless going down you are seeing the error message on the player. Sorry about that.

I don't think it is player type related then as SB3, TR, SBR and SBB use the same mechanism / code to download new firmware.

Could it be that the load on a (weak) wireless network is too high if a bunch of players start downloading firmware at the same time?

If this is the case then there isn't much firmware could do I think, but maybe SC / SN could 'schedule' firmware upgrades and ensure not all happen at the same time?

QA: I would like to change the test then to multiple Booms connected wireless to one SC.
Comment 9 Max Spicer 2008-10-20 04:36:27 UTC
My wireless network isn't _that_ bad - it can sustain some pretty large transfers without ever freezing.  It only freezes when this fw upgrade error occurs.  No idea why.  I guess I should try eliminating devices, but time is pretty limited at the moment unfortunately.
Comment 10 Felix Mueller 2008-11-04 03:43:41 UTC
Moving out bug that won't make it for 7.3
Comment 11 Felix Mueller 2008-11-06 09:32:13 UTC
Doh, there will be no 7.4 - moving it to 8.0
Comment 12 Felix Mueller 2008-12-15 10:20:29 UTC
Maybe related. Seems like the users player had its MAC address reverted to the factory MAC address and after correcting the MAC address the firmware update worked again.
Comment 13 Sean Adams 2008-12-15 13:21:29 UTC
It's a long shot, but the recently fixed TCP reassembly bug might explain this. If we have a test case it would be good to try it with that change.
Comment 14 James Richardson 2009-10-05 14:30:39 UTC
This bug has been marked as fixed in the 7.4.0 release version of SqueezeBox Server!
    * SqueezeCenter: 28672
    * Squeezebox 2 and 3: 130
    * Transporter: 80
    * Receiver: 65
    * Boom: 50
    * Controller: 7790
    * Radio: 7790  

Please see the Release Notes for all the details: http://wiki.slimdevices.com/index.php/Release_Notes

If you haven't already, please download and install the new version from http://www.logitechsqueezebox.com/support/download-squeezebox-server.html

If you are still experiencing this problem, feel free to reopen the bug with your new comments and we'll have another look.