Quote:
Originally Posted by demonstar55
This is happening on PEQ
EDIT: Akkadius also reported this happening on EZ, but he didn't realize he did :P (the raid window doubling up toons is a symptom of whatever is going on) I'm also pretty sure it's not limited to RoF2
|
I don't doubt there are issues with raids like that which still need to get corrected. Since normal groups and raid groups are handled mostly separately in the source, an issue seen with raids won't necessarily exist in normal groups.
I believe I have worked out the majority of the group issues. I haven't touched raid stuff at all. The last issue I recall is there was a bad query related to raid leaders, which is the same query that was failing for groups that I fixed a while back. I don't recall if that raid leaders query ever got fixed, but I am pretty sure it could cause issues related to zoning and raids. IIRC, it is a simple fix to just change an INSERT to a REPLACE INTO in the query for the raid leaders table updates.
I too agree that this is unlikely to be an RoF2-specific issue. It probably happens on all clients. If it isn't the raid leaders query issue, then I wouldn't even know where to start troubleshooting it unless someone can reproduce it 100% and provide those steps here.
For any server admins that see this issue, I would first check your related tables to see what is showing in there. I have seen some group related issues popup a few times over the past few months related to the auto-update script. This would only be an issue on Linux servers that have a persist_world script and what can happen is the update script prevents world from starting if you have required updates to make and don't go through the menu. But, then the persist_world script launches and starts restarting world every 2 seconds. When world starts, part of the process is to delete all groups from the database, so basically all groups poof every 2 seconds. If you zone in this scenario, the group will be gone every time. To fix it, you have to kill the script process and world process, then start world by itself (without the start script) and run through the menu to update the required updates. Then, kill world again and start your server normally with the start script. The permanent fix is to add the update script into your startup script prior to starting any of the actual server processes. This way, by the time world is started, the updates are already all completed.
The above is probably not what is causing the reported issues in this thread, but something worth noting in case anyone runs into it.