Did a test import of 1000 songs into the beta Ok.
Selecting 1000 songs to auto-label.
At the 3th set of 100 from the 1000songs. I noticed that the progress was not updated in the header.
(3th set is 201 to 300)
See image i took of the 3th set.
https://imgur.com/a/wGOSPLC
Also the dutch language is off there, the word "naar" translate in this context to something like "go to"
Auto-label progress not updated in header And white screen crash
Moderator: Gurus
-
CrashMonkey
- Posts: 45
- Joined: Sat Oct 21, 2023 10:34 am
Auto-label progress not updated in header And white screen crash
Last edited by CrashMonkey on Tue Oct 24, 2023 5:57 pm, edited 1 time in total.
-
CrashMonkey
- Posts: 45
- Joined: Sat Oct 21, 2023 10:34 am
Re: Auto-label progress not updated in header
And while labeling in the 3th set. the program froze to a white screen. Needed to exit it trough taskmanger.
https://imgur.com/XTUylO5
When restarting 5.1 and looking in the log file it showed the current session. And not the one with crash.
However i did run the debugview
and when screen froze this happend at same time:
I will upload upload the debug.log to a ticket.
Edit: [Ticket #6425]
https://imgur.com/XTUylO5
When restarting 5.1 and looking in the log file it showed the current session. And not the one with crash.
However i did run the debugview
and when screen froze this happend at same time:
Code: Select all
00628213 6490.55126953 [12424]
00628214 6490.55126953 [12424] <--- Last few GCs --->
00628215 6490.55126953 [12424]
00628216 6490.55126953 [12424] [12424:40C14000] 6477001 ms: Mark-Compact 972.4 (1042.2) -> 959.1 (1040.2) MB, 879.4 / 0.1 ms (average mu = 0.342, current mu = 0.367) allocation failure; scavenge might not succeed
00628217 6490.55126953 [12424] [12424:40C14000] 6477216 ms: Scavenge 966.5 (1040.2) -> 962.1 (1040.2) MB, 6.8 / 0.0 ms (average mu = 0.342, current mu = 0.367) task;
00628218 6490.55126953 [12424] [12424:40C14000] 6477397 ms: Scavenge 968.7 (1040.2) -> 964.5 (1042.2) MB, 9.7 / 0.0 ms (average mu = 0.342, current mu = 0.367) allocation failure;
00628219 6490.55126953 [12424]
00628220 6490.55126953 [12424]
00628221 6490.55126953 [12424] <--- JS stacktrace --->
00628222 6490.55126953 [12424] I will upload upload the debug.log to a ticket.
Edit: [Ticket #6425]
Last edited by CrashMonkey on Thu Nov 02, 2023 4:40 pm, edited 2 times in total.
-
CrashMonkey
- Posts: 45
- Joined: Sat Oct 21, 2023 10:34 am
Re: Auto-label progress not updated in header
Did a second run with the exact the same songs, because something strange happend during the 3th batch in first run.
Explorer.exe crashed suddenly. WhoCrashed told me that destructor'.sys was the culprint.
I guess somewhere is destructor called but then in code as 'destructor'' <-- one quite --> ' <-- to much? (both crashes caused by the same thing)
To be sure if it was mediamonkey, i did the re-run.
At the 3th set of 201 of 300 from 1000, it made it crash again around track 28. This happend at the first run at around the same track.
When i read the GC failed, i read it as Garbage Collector? I am i right?
Then i guess the instance before that the GC failed maybe to destuct the items, resulting in my explorer crash, and the next time it ran, it could not allocate because it was not destroyed before? Resulting on white screen of death.
The two above lines of text are a guess.
I did this run, without the debugview. Without debug view autolabel header does update ok. Tomorrow a run again with the debugview confirming once again if all turns out the same.
One more annoyance, when its autolabeling is going. And while i write this, with every track change, the focus goes off my google chrome for 1 sec, and i cannot type, then focus comes back and i can type again. This happens also when i go to a other application.
And while i write this the focus went out of my browser back to MM, and it crached with a white screen. On the end of the 3th run again.
Will do this tomorrow again, and i will install a video recording tool, to demonstrate it in a video. (As for a other strange behaiver i might need it as well, to show whats going on. But that occurs on a more rare basis)
Edit: In version 5 of MM, the behaivor is that the entire app crashes. Thats why i am testing it now in 5.1
Explorer.exe crashed suddenly. WhoCrashed told me that destructor'.sys was the culprint.
I guess somewhere is destructor called but then in code as 'destructor'' <-- one quite --> ' <-- to much? (both crashes caused by the same thing)
To be sure if it was mediamonkey, i did the re-run.
At the 3th set of 201 of 300 from 1000, it made it crash again around track 28. This happend at the first run at around the same track.
When i read the GC failed, i read it as Garbage Collector? I am i right?
Then i guess the instance before that the GC failed maybe to destuct the items, resulting in my explorer crash, and the next time it ran, it could not allocate because it was not destroyed before? Resulting on white screen of death.
The two above lines of text are a guess.
I did this run, without the debugview. Without debug view autolabel header does update ok. Tomorrow a run again with the debugview confirming once again if all turns out the same.
One more annoyance, when its autolabeling is going. And while i write this, with every track change, the focus goes off my google chrome for 1 sec, and i cannot type, then focus comes back and i can type again. This happens also when i go to a other application.
And while i write this the focus went out of my browser back to MM, and it crached with a white screen. On the end of the 3th run again.
Will do this tomorrow again, and i will install a video recording tool, to demonstrate it in a video. (As for a other strange behaiver i might need it as well, to show whats going on. But that occurs on a more rare basis)
Edit: In version 5 of MM, the behaivor is that the entire app crashes. Thats why i am testing it now in 5.1
-
CrashMonkey
- Posts: 45
- Joined: Sat Oct 21, 2023 10:34 am
Re: Auto-label progress not updated in header And white screen crash
Update on "Auto-label progress not updated in header"
Have tested this a few times now, the header does not update when DebugView.exe is started before MM is started.
So the header does update OK, when running MM normaly. Only when debugging it won't update.
Have tested this a few times now, the header does not update when DebugView.exe is started before MM is started.
So the header does update OK, when running MM normaly. Only when debugging it won't update.
-
CrashMonkey
- Posts: 45
- Joined: Sat Oct 21, 2023 10:34 am
Re: Auto-label progress not updated in header And white screen crash
Have tested this today again in the latest beta 5.1.0.2821
Test case is auto-labeling 1000 song's that where never managed before. (1000 same song's i restore from backup every new attempt)
Auto-labeling seems to work beter then in the last version, but still have issues.
- Auto-label progress not updated when connected with debugview is working. So solved OK
- In all other version i tested, auto-labeling always crashed somewhere between 250 and 350 song. For the first time i was able to go around 850 of the 1000.
However i saw 3 times this error popup, maybe more as i did not continue to watch the screen. viewtopic.php?t=100545
I have updated the ticket 6432, with full debugview logfile of today.
Edit: Every time i do a big test like this, my google account get logged out and need to reverify my account? I guess it has to do with to many connections made within a certain time for illustration lookup. Is this known to happen?
Test case is auto-labeling 1000 song's that where never managed before. (1000 same song's i restore from backup every new attempt)
Auto-labeling seems to work beter then in the last version, but still have issues.
- Auto-label progress not updated when connected with debugview is working. So solved OK
- In all other version i tested, auto-labeling always crashed somewhere between 250 and 350 song. For the first time i was able to go around 850 of the 1000.
However i saw 3 times this error popup, maybe more as i did not continue to watch the screen. viewtopic.php?t=100545
I have updated the ticket 6432, with full debugview logfile of today.
Edit: Every time i do a big test like this, my google account get logged out and need to reverify my account? I guess it has to do with to many connections made within a certain time for illustration lookup. Is this known to happen?