So, system Eos Classic Main (in the control room, untouched); Eos Ti Backup (in the theatre, active control surface), Eos RVI (LD monitors). All running 2.9.0 build 77. Drama school show with students operating.
This system was fine and stable all day yesterday.
Today we re-organised our displays and recorded that setup as an 'all workspaces' snapshot.
From that point on, if we go into blind three times in quick succession (ie. blind-live-blind-live-blind-crash), the Ti crashes to the shell. Other consoles unaffected.
When the Ti re-starts and re-connects it is no longer in the snapshot.
If we recall the snapshot (snapshot 21), the displays come back properly but blind-live-blind etc then crashes as above. This is completely repeatable, though I can't say for sure it's always three times - sometimes two, sometimes four, so it might be time related rather than number of display changes related.
BUT if we re-start and re-connect then DON'T recall the snapshot, everything seems to work fine - I can go blind-live as often as I want and nothing bad happens.
Only 'oddity' I can see is that the LD is the same user number as the Ti on the RVI, but only has one display connected and has this set up differently, so it's slightly unclear what is being recorded into the snapshot.
Actually, one other oddity: before we made the snapshot, on one of our external monitors the 'close tab' option was greyed out on many of our tabs, even though the tab frame was not locked. Eventually I 'closed all but this tab' then re-built the displays before saving the snapshot.
Would be good to know if anyone spots anything odd...
Logs and showfile below.
Esmeralda - Tuesday 11th Feb 2020-02-11 09-56-36.esf
I was unable to re-create this crash using just a primary. I then went on to give it a shot from a backup console—still nothing. Then, I tried with your show file (but without recalling any snapshots) and could not get it to crash. It wasn't until rapidly swapping between live and blind AFTER recalling snapshot 21 that I was able to get it to crash.
Could you please re-send your Backup's logs (this time, going to Advanced>Include Large Crashes)?
Also, could we get logs from the Primary as well?
I'll be messaging you a link that you can upload them to shortly.
Can't get to the main right now (it's locked in a control room, I don't have the key!).
But: if you look at my second message and follow that you might be able to reproduce the crash without running the snapshot. I've been going '1 THRU 600 VIEW CHANS'. The crash seems to happen if tab 1 live is in flex-all and tab 1 blind is in 1 THRU 600 FLEXI VIEW CHANS, or vice versa. If both are in view chans or neither then it doesn't crash.
(Snapshot 21 recalls tab-1 with live set to flexi all and blind set to flexi view-chans; that's why it's triggering this, rather than the snapshot itself being the cause).
Let me know if you still need the master's logs and I'll try to break in to the room!! ;-)
Still having trouble getting my system to crash without firing that snapshot.
Using your show file, I've got Live set to no Flexi and Blind set to View Chans 1-600. If I rapidly switch between live and blind, I can't crash it. After firing snapshot 21, I can crash it easily.
If you can at least pull logs from the Backup again with Large Crashes included, we can get started with those if you don't have access to the primary right now.
Will grab them next meal break.
But I can definitely do this consistently on the Ti backup:
- run snapshot 21, blind-live-blind- console crashes (sometimes one more -live-blind)
- restart the console; re-starts as if snapshot 21 had NOT been run. DO NOT run snapshot 21.
- tab1, blind, 1 thru 600 FLEXI VIEWCHANS
- blind-live- console crashes (usually first blind, sometimes second)
- start the console
- snapshot 21
- blind, tab-1, flexi-all (so no longer in flex-viewchans)
- blind-live-blind- repeat - no crashing
- tab 1, in live, 1 thru 600 FLEX VIEW CHANS (so now both live and blind are in flexi viewchans)
- blind-live-blind-live-repeat - no crashing
Logs from the main (Eos Classic) and logs including big crashes from the Ti (backup).
Thanks Rob! These have been submitted for analysis