Mozilla DB News, Fri 31 August

  • Reduced the amount of noise being sent from cron so that we would actually be looking at the e-mails we get.
  • Set up Tokutek for our Datazilla program.
  • Debugged and solved a problem with the Tokutek engineers because tokudb’s mysqld was not recognizing the –defaults-file option (it’s a MySQL quirk, –defaults-file has to be the first option).
  • Cleaned up and fixed a bunch of MySQL ACLs on several servers.
  • Debugged a problem where slow query logs were not being copied to a central location. The solution involves changing the script to have splay, ssh keys, and opening some network flows.
  • Changed Nagios alerts so that dev/stage machines only page during the workday, not during weekends or evenings.
  • Added some machines to Graphs.
  • Set it so when the Crash-stats staging database is refreshed from production, only the reload can connect to it while the database is loading.
  • Fixed a replication problem that occurred because the slave had MyISAM tables but the master had InnoDB tables.
  • Re-synchronized 2 separate machines after a disk failure caused InnoDB corruption.
  • Added new permissions to Bugzilla due to schema changes.
  • Deleted some spam from a Bugzilla bug.

  • Reduced the amount of noise being sent from cron so that we would actually be looking at the e-mails we get.
  • Set up Tokutek for our Datazilla program.
  • Debugged and solved a problem with the Tokutek engineers because tokudb’s mysqld was not recognizing the –defaults-file option (it’s a MySQL quirk, –defaults-file has to be the first option).
  • Cleaned up and fixed a bunch of MySQL ACLs on several servers.
  • Debugged a problem where slow query logs were not being copied to a central location. The solution involves changing the script to have splay, ssh keys, and opening some network flows.
  • Changed Nagios alerts so that dev/stage machines only page during the workday, not during weekends or evenings.
  • Added some machines to Graphs.
  • Set it so when the Crash-stats staging database is refreshed from production, only the reload can connect to it while the database is loading.
  • Fixed a replication problem that occurred because the slave had MyISAM tables but the master had InnoDB tables.
  • Re-synchronized 2 separate machines after a disk failure caused InnoDB corruption.
  • Added new permissions to Bugzilla due to schema changes.
  • Deleted some spam from a Bugzilla bug.