Hey, guys, I'm groovin' on the new server. Cheers!
So, the past couple days I've been having my BeanCounter.lua empty out on me. Poked around, saw I'm getting the "constant table overflow". Did some digging, and discovered I'm up against the .lua limit, because, naturally, I've never purged any BeanCounter data. So, I restore from backup, shift-click my "Enable purging transactions..." (set to 48), and...nada. I hit "Done"...bupkis. I log out and back in...all my data are still there. "Well, I know I've got older data than 48 months," I say to myself, "but, let's just really clean it out." I set the slider to 6 months, shift-click "Enable" again, hit "Done" again, relog again, and...goose egg. I've still got over 200,000 Entries, and data going back to at least 10/09.
What's the magic command, guys? Where's the flush handle on this sucker? The wiki's got nothing, and my feeble searches in Discussions have yielded no joy.
If the option is enabled then it will purge old data the next time the player logs into a character and performs a manual search.
The first manual search of each session is when most of the "data maintenance" functions actually run. They are currently only run per toon, not account wide so each toon that has data will need to to be triggered.
Kandoko did post this while helping me, perhaps you will find it useful too:
------ You can double check by typing /run print(BeanCounter.GetSetting("oldDataExpireEnabled") ) and /run print(BeanCounter.GetSetting("monthstokeepdata") )
That should print "true" and # of months of data before it gets removed
You can force a run of the cleaning. Log into each toon and In chat type /run BeanCounter.Private.compactDB(true) -------
It looks like that the settings are global, but you MUST run the "/run BeanCounter.Private.compactDB(true)" command on each toon that has the data in the Beancounter. After the compact do a /reload and the Entry count should match the current amount of data you have.
Yes, this has been changed since zab42 info. The current system works this way.
Each toon has a maintenance schedule. Each login Beancounter looks at this schedule and runs each toons task.
So there is no longer a need to login to a server/character to get the database purges to work. szynaka code will make the purge happen immediately for the "Server" and "Character" names you enter in quotes
AnrDaemon, running every task for every toon at on the same login was barely noticeable on the 10 MB multicharacter DB.
The tasks have a fixed value + random padding added each time.
Each toon and each task has its own schedule. . The tasks that could be run for character include, "sortArray" schedule to run monthly + 1 to 5 day random drift
"prunePostedDB" schedule to run every 14 days + 1 to 5 day random drift
"compactDB" schedule to run monthly + 1 to 5 day random drift
All of these "tasks" were running for every server and every toon at every login for a few months in DEV before I put them on a schedule btw.
Though so, but had to ask anyway. Looking into login slowdowns, but so far can't pinpoint exact cause :( Probably an interference, though, but I can't get it clear yet...
Nah, add some print statements to the tidyup code. I could have a bug making them run too much or something. I was rushing to get that stuff finished before my RL stuff took over.