I say “for now” because this is the same architecture that ultimately failed, when the transaction logs filled up the storage. Bigger storage this time, but the clock’s still ticking. I’ll have to return to the db problem after I get my current (business) project finished.

Go ahead and resume posting, but be aware that there’s a non-zero risk that the database will just as capriciously stop working again. You might lose a post until I can replace the database server entirely and copy over the data. That won’t be tonight. It’s late.

There was a conversation here for a while, during a phase when the DB was (also anomalously) working, you guys speculating on the problem. I think it was mcfly who reminisced about the damn transaction log files, and indeed, it was the damn binlogs. I had three databases arranged as a master and two slaves, and the binlogs recorded every change to the database as they were propagated out from the master to the slaves. The master hard drive filled up with binlog, database files were corrupted, and the house of cards fell down.

Clearly I have some redesign work to do as I rebuild a new master and slave servers.

Sorry you guys went without your Zone for a whole day. But I hope you got outdoors and aired yourselves out a bit. Got rid of that musty internet smell.