Hey Staff team, thanks for taking a look at my post . I run Eroded Networks, hosting Plutonium Zombies servers for the AU community. I’d really appreciate your help with server-browser visibility and persistent hitches/hangs.
My AU setup:
• Sydney datacentre, Faded Servers
• i7-10700K, 8 cores/16 threads, 64 GB RAM
• Ubuntu 24.04.4 LTS, kernel 6.8.0-142
• Pterodactyl 1.15.1, Wings 1.13.1, Docker 29.7.1
• Wine 9.0, win64 prefix, launched through xvfb-run inside Docker
• AU BO2 Reimagined currently reports Plutonium r5354
• Custom GSC and native utility integrations, which I’m also investigating
SERVER BROWSER
Players get good ping and can connect directly by IP, but servers can take an extremely long time to appear in the browser. Sometimes individual servers never appear, despite remaining accessible. From the checks I’m watching, they aren’t missing heartbeat check-ins.
The hitches and stability problems affect all my BO1 and BO2 servers, including Official and Modded, not just Reimagined. The Buried incident on Reimagined is the latest example, but I need help investigating the wider issue across both games.
What exactly does a successful heartbeat confirm? Can registration succeed but a separate query/discovery check prevent listing? Are there known issues with multiple servers on one public IP, query ports, rate limits, browser caching or AU routing?
What timestamped logs and server details would help you trace registration versus browser discovery from your side?
HITCHES / HANGS
I’m getting hitches during gameplay and occasional servers that remain running but stop responding. Checks and work so far include:
• Correlating game logs/GSC progress with UDP queries, game-thread CPU, memory pressure and disk I/O.
• Investigating disk-latency spikes, MariaDB reads and replay-checkpoint writes. Checkpoint writes were reduced, the DB buffer pool increased to 2 GiB, and buffered BO2 logging verified.
• Separating loading/script-init warnings from populated gameplay stalls. Some gameplay cases had no matching DB read storm.
• Investigating GSC pools. One earlier BO2 Modded incident reached 65,536/65,536 child variables, 17,325 parents and 2,148 threads. That doesn’t establish the cause of every hitch.
• Testing delayed HTTP responses through the utility DLL. Requests serialised and a GSC timeout didn’t cancel the native request, but game progress continued; it did not reproduce a main-thread freeze.
• Investigating BO1 processes that stayed alive/bound but stopped answering queries, without matching OOM evidence.
Are there recommended Wine versions/settings for T4/T5/T6? What’s the best supported way to capture the offending game thread/GSC function, inspect pool/thread pressure and check native-plugin compatibility?
I haven’t completed a controlled native-Windows comparison, so I’m not claiming Wine, host contention or my scripts are ruled out. What controlled test would you prioritise?
Also, could I get the Server Hoster role? Please let me know what verification you need.
I’m putting a lot into keeping reliable AU servers available and would genuinely appreciate your guidance. Cheers!