----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Handle millions of tables (add dictionary) CREATION DATE..: Thu, 30 Jun 2011, 17:48 SUPERVISOR.....: IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 223 (http://askmonty.org/worklog/?tid=223) VERSION........: WorkLog-4.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 200 (hours remain) ORIG. ESTIMATE.: 200 PROGRESS NOTES: DESCRIPTION: Information_schema falls apart when there is millions of tables. Fix is to add a table dictionary of the existing files into a mysql.dict file and use this instead of the file system to return information schema queries that doesn't require a table to be opened. When we try to access a table that doesn't exists in the dictionary, we should open it to see if the .frm file exists and if yes, add it to the dictionary. In a similar fashion, if a table that should exist doesn't, we should remove it from the dictionary. We would also need a refresh command to update the dict from the .frm file (to be used if one copies a table over an old table with different information). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v4.0.0)