im hoping its one of the issues we are aware, is there any other info I can provide to help track down the issue?
I have taken a look at the mdmp files, the main dayz thread is getting stuck on a WaitForSingleObject but with debug symbols thats about as far as I can go.
We have also reduced the max player count from 105 to 90, this seems to have eased the issue but not resolved it completely.It is happening multiple times, normally during peak server time.
It was happening between 4pm - 8pm (peak times), around 6 hangs over that time period.
Occurrence count: at least 24 separate freeze/hang events, based on logs retained from Aug 7 onward. Two distinct phases:Phase 1 — sporadic freezes, Aug 7 12:30 through Aug 8 18:58 (16 occurrences)
Freeze-to-restart intervals ranged 12–130 minutes, no consistent rhythm:Aug 7 12:30:27 → froze 13:56:47 (86 min)
Aug 7 14:35:51 → froze 15:08:46 (33 min)
Aug 7 15:46:32 → froze 16:19:54 (33 min)
Aug 7 16:47:06 → froze 18:49:05 (122 min)
Aug 7 19:26:37 → froze 20:00:43 (34 min)
Aug 7 20:37:37 → froze 20:49:33 (12 min)
Aug 8 04:30:17 → froze 05:51:24 (81 min)
Aug 8 06:27:12 → froze 07:31:22 (64 min)
Aug 8 07:56:12 → froze 08:20:46 (25 min)
Aug 8 08:56:17 → froze 09:58:48 (63 min)
Aug 8 10:20:13 → froze 10:50:55 (31 min)
Aug 8 11:42:52 → froze 12:08:34 (26 min)
Aug 8 12:30:17 → froze 13:20:55 (51 min)
Aug 8 13:50:19 → froze 15:21:07 (91 min)
Aug 8 15:43:53 → froze 17:53:34 (130 min)
Aug 8 18:15:47 → froze 18:58:02 (42 min)
Phase 2 — tight, consistent loop, Aug 9 08:30 onward (8 occurrences, ongoing)
Every restart since 08:30 has frozen in an 11–18 minute band with an identical terminal crash signature (Unhandled exception — Access violation, illegal write by 0x7ff75e114ca0 at 0x0):08:30:17 → froze 08:48:27 (18 min)
08:53:17 → froze 09:09:04 (16 min)
09:13:52 → froze 09:26:04 (12 min)
09:30:52 → froze 09:44:38 (14 min)
09:49:27 → froze 10:00:50 (11 min)
10:05:43 → froze 10:21:54 (16 min)
10:26:42 → froze 10:42:25 (16 min)
10:47:13 → froze 11:00:43 (13 min)
All times are server-local (UTC-7). Freezes are confirmed via the
freezeCheckwatchdog snapshotting the still-running-but-unresponsive process (FROZEN_*.mdmp), followed several minutes later by a forced termination. Note: log retention has a gap (Aug 2–Aug 6) so earlier occurrences outside this window can't be ruled out or confirmed.