Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Looking forward to the next drop (6.x?)
#31
So even 15 is just a value update pass, i.e. going through the widgets to find the ones that are field/variable based and giving them a chance to update their state. That's not the one that I would have expected. See if you can get another one of those just to make sure that this isn't just random. If it's the same event again, then presumably it really is somehow managing to get hung up during that which would be awfully odd.
Dean Roddey
Explorans limites defectum
Reply
#32
9:09 PM last night all iPads Frozen.
Nothing in the log.
After I restarted one iPad this showed up in log.

05/31 06:53:43-CQC-PC, CQCWebSrv, WebSockHandThreadThread41
{
    CIDLib, CIDLib_Thread.cpp.1320, Failed/Timeout, Error: 2433/13/0
    Timed out waiting for operation to complete on thread 'CQCWebRIVAGUIThread_3'
      <CQCWebSrv> CQCWebSrvC_WebRIVAHandler.cpp - 1240

}
05/31 06:53:43-CQC-PC, CQCWebSrv, WebSockHandThreadThread41
{
    CQCWebSrvC, CQCWebSrvC_WebRIVAHandler.cpp.1246, Status/Termination
    The Faux GUI thread could not be stopped. Event=15
}
_______________
Denon 3808ci, 2112ci , SONOS, Russound CAA66,Z-Wave(Lights,Locks), Hue, SmartThings,
iPads,Tivo,Hikvision,Elk-M1,iTunes,TED5000,Somfy RTS blinds, Amazon Echo, Google Home, Polk XRT12,
Honeywell Wi-Fi 9000 & Caleo Wi-Fi Thermostats, Rainmachine
Reply
#33
Same event, so presumably it's for real. That's a bit of a mystery. I'll have to come up with some way to debug that.

In the meantime, though I don't expect anything to come of it, let's continue the divide and conquer with an easy test and just have you bring up the same template on Chrome or Edge or something and leave it running. Just to continue replacing each moving part to see what happens. Though it's unlikely to matter, it's so easy to do that it's worth it just to be sure.
Dean Roddey
Explorans limites defectum
Reply
#34
Oh, something I completely forgot about, if you add this:

logguievents=1

to the URL, it will cause the WebServer to log all of the faux GUI thread events it is processing. So it would be worth doing that. Only do it on one client, since it will generate a lot of output and we don't want to have to dig through that times X and find the events for a particular client.

Normally this might be an issue since it might have cycled the original problem out of the logs before you notice it's down and get the dump. But, in this case, since the thread that is logging them apparently gets stuck, it should stop logging at the point where it locks up and we can see what has been happening up until then.
Dean Roddey
Explorans limites defectum
Reply
#35
Actually, wait for for me to refresh again, since I'm going to add some more useful logging to help diagnose the issue.
Dean Roddey
Explorans limites defectum
Reply
#36
OK, I refreshed 957 again. So this should log all of the non-repetitive GUI events (it doesn't log field/variable update passes since they are constant.) So we should have a good idea of what led up to the issue. Just set the logguievents=1 value to one of the clients. It would be useful also to set the sessname=whatever value as well to make it easy to find all msgs from that client.
Dean Roddey
Explorans limites defectum
Reply
#37
(05-31-2019, 12:09 PM)Dean Roddey Wrote: OK, I refreshed 957 again. So this should log all of the non-repetitive GUI events (it doesn't log field/variable update passes since they are constant.) So we should have a good idea of what led up to the issue. Just set the logguievents=1 value to one of the clients. It would be useful also to set the sessname=whatever value as well to make it easy to find all msgs from that client.

OK,  all iPads froze last night at 11:51pm  .  Both safari and Chrome.
Nothing in log around that time.
new items in log are from this morning when I restarted one client.
Thanks
Kevin

06/05 20:59:15-CQC-PC, CQCWebSrv, CQCWebRIVAGUIThread_2
{
    CQCWebSrvC, CQCWebSrvC_WebRIVAHandler.cpp.510, Status/App Status
    [Unknown] Processing press event
}
06/05 20:59:15-CQC-PC, CQCWebSrv, CQCWebRIVAGUIThread_2
{
    CQCWebSrvC, CQCWebSrvC_WebRIVAHandler.cpp.510, Status/App Status
    [Unknown] Processing release event
}
06/06 07:28:17-CQC-PC, CQCWebSrv, WebSockHandThreadThread21
{
    CIDLib, CIDLib_Thread.cpp.1320, Failed/Timeout, Error: 2433/13/0
    Timed out waiting for operation to complete on thread 'CQCWebRIVAGUIThread_2'
      <CQCWebSrv> CQCWebSrvC_WebRIVAHandler.cpp - 1248

}
06/06 07:28:17-CQC-PC, CQCWebSrv, WebSockHandThreadThread21
{
    CQCWebSrvC, CQCWebSrvC_WebRIVAHandler.cpp.1254, Status/Termination
    The Faux GUI thread could not be stopped. Event=15
}
_______________
Denon 3808ci, 2112ci , SONOS, Russound CAA66,Z-Wave(Lights,Locks), Hue, SmartThings,
iPads,Tivo,Hikvision,Elk-M1,iTunes,TED5000,Somfy RTS blinds, Amazon Echo, Google Home, Polk XRT12,
Honeywell Wi-Fi 9000 & Caleo Wi-Fi Thermostats, Rainmachine
Reply
#38
OK. There was nothing between the last button press and the failure a couple hours later. So that would pretty much mean that one of the many and ongoing update passes just gets hung for some reason. That's pretty strange. It would have to be something different about that ultimate failing pass. Either some resource exhausted or something else just happens to happen at the same time and it causes some sort of deadlock or something. But, if it were a deadlock, one of the other threads involved would also not be stoppable, and there doesn't seem to be any evidence of that.

This is going to be hard to diagnose.
Dean Roddey
Explorans limites defectum
Reply
#39
OK, it looks like the problem was constantly re-registering fields in the poll callback. That's not something that should really be done and though everything should in theory survive, it's not something I'm going to put a lot of time into figuring out right now, i.e. it's not something that should hold up a release. I was waiting because I was worried it was some subtle side effect of all the code changes of late. But apparently not, so probably safe enough to move forward.
Dean Roddey
Explorans limites defectum
Reply


Possibly Related Threads...
Thread Author Replies Views Last Post
  Text to Speech changes in 5.0.9 and forward Dean Roddey 0 2,014 01-12-2017, 06:18 PM
Last Post: Dean Roddey
  Audio Drop Out simon 3 2,352 07-28-2007, 12:24 PM
Last Post: simon
  Next drop date simon 1 2,070 04-22-2007, 11:21 AM
Last Post: Dean Roddey

Forum Jump:


Users browsing this thread: 1 Guest(s)