Site Freezing and Odd Behavior

[escorpius]

What's in the Error Log during these crashes? Do you have the KillInactiveSessions markers recording and if so are any crashes preventing the End marker from recording.

The latest does show a fail after KillInactiveSessions Begin but before KillInactiveSessions End. See the excerpt here, which shows the results that span the time from the freeze through the restart:

Code: Select all

2022-10-19 19:23:14|[debug] KillInactiveSessions Begin
2022-10-19 19:24:55|Unexpected balance (121) in House Ring account at startup
2022-10-19 19:24:55|Crash recovery implemented, details in Event Log
2022-10-19 19:25:56|[debug] KillInactiveSessions Begin
2022-10-19 19:25:56|[debug] KillInactiveSessions End

DIsconnects are generated way up the chain at the O/S level. The software just reports what it sees (unless it's manually killing the connection itself).

More a question on whether or not there are specific settings on server-side (or client-side) that could help or other things you've seen that could cause these issues. I note there are other threads in the forum that reference this behavior. We've not seen it at this level in the past, so hard to dismiss the likelihood that it's related, somehow.

There's no way for the poker software to distinguish the cause.

From the server side, I understand that a disconnect looks like a disconnect. Was wondering if anything is done client-side that could help diagnose, e.g., console logging that could be viewed through developer tools.

[Kent Briggs]

From the server side, I understand that a disconnect looks like a disconnect. Was wondering if anything is done client-side that could help diagnose, e.g., console logging that could be viewed through developer tools.

The client creates a WebSocket object in javascript and monitors its onclose event where it will dump the event code to the debug window (which the player can see by long pressing the Help menu):

Code: Select all

debug("MSG Connection Closed with Event Code " + e.code);

But I doubt it will tell you anything useful. I think most browsers just use a basic disconnect code.

[escorpius]

We'll give it a try, thanks!

[escorpius]

We are still seeing relatively frequent crashes. We tend to go several days before they happen, but have seen 3 in the last 2 days. The only output lines in the ErrorLog consistently look like this at the moment of the crash:

Code: Select all

2022-10-27 19:34:34|[debug] KillInactiveSessions End
2022-10-27 19:35:34|[debug] KillInactiveSessions Begin

The tail of the latest crash EventLog looks like this:

Code: Select all

2022-10-27 19:35:46|Connect|yhgg disconnects session 00000054, PC 73BDFFAA
2022-10-27 19:35:47|Connect|Connection 00000067 accepted from IP ***.***.72.7, PC 73BDFFAA
2022-10-27 19:35:54|Connect|IP ***.***.72.7 disconnects session 00000067, PC 73BDFFAA
2022-10-27 19:35:57|API|AccountsSessionKey from ***.***.231.235 (yhgg)
2022-10-27 19:36:01|Connect|Connection 00000068 accepted from IP ***.***.72.7, PC 73BDFFAA
2022-10-27 19:36:02|Logout|yhgg logged out session 00000054, PC 73BDFFAA
2022-10-27 19:36:02|Login|yhgg logged into session 00000068 from IP ***.***.72.7, PC 73BDFFAA
2022-10-27 19:36:04|Connect|M-1 disconnects session 00000062, PC 20702AC7

[BlackKnite69]

FWIW, The crash , each time happens is pairs. The site freezes, we have software that detects the crash and gets the site back up online. within seconds of players signing in, the site crashes a second time. Once it is back online the second time, it remains up and running, for some random length of time.

[Kent Briggs]

We are still seeing relatively frequent crashes.

Are you still monitoring the PM_Timer thread and is it still freezing?

[AdelinaMoore]

I analyze all reports on this theme. That no software problem, like as WinSAT load you system to 100% that do freeze you system on random time 1-2 sec, this may be freeze PMService. Analyze you OS system on process when have freeze and CPU load too, like a process hacker log. And log ping port inside ASW on loop (ping youself) - its may be get info about when freeze work on adapter. You say - use AWS. Check NAT network inside AWS and recreate you VM to clear point. Next point check rules to aceess dbfile. If AWS have proactive system they can freeze 1-2sec on inspect=its Virtualized RAM settings on you OS Windows In Windows Defender its Memory Integrity - Disable it = https://support.microsoft.com/en-us/win ... yintegrity

About run method - Check how user do Service PMSevice - run that service on Administrator.