Switch to upstream backup and restore scripts
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
mythtv (Ubuntu) |
Fix Released
|
Medium
|
Unassigned |
Bug Description
Binary package hint: mythtv
<sphery> superm1: I heard you guys are using a *buntu-specific backup script. What do you think of adding --add-drop-table to the mysqldump options? Doing so should prevent the issues users are having when they restore a 0.21 schema version backup on top of a 0.22 schema database (the blank one installed by the packages)--which means they end up with 0.22 tables (because of the CREATE TABLE IF NOT EXISTS) with 0.21 data. (Or if you guys ...
<sphery> ... have a restore script, fix that so it checks to ensure the DB is empty before allowing the restore, like the one in programs/
<sphery> I'm considering adding --add-drop-table to the programs/
<sphery> that way, even if they don't use the restore script, it will prevent issues
* rotorr has quit (Read error: 104 (Connection reset by peer))
* rotorr (<email address hidden>) has joined #mythtv
* Prost (<email address hidden>) has joined #mythtv
<superm1> sphery, well I think we are better switching to the upstream backup and restore scripts tbh
<superm1> it's a needless delta currently
<sphery> that would work, too... My main concern is getting the word out to users--I've seen this happen 4 times, now, and 2 of 3 hacked the schema 'til the upgrade completed, which means their data and schema is likely suspect)
Related branches
Changed in mythtv (Ubuntu): | |
importance: | Undecided → Medium |
milestone: | none → lucid-alpha-1 |
status: | New → Triaged |
This has been fixed in the trunk packaging. It will be seen in the next 0.23 autobuilds trunk build available at http:// www.mythbuntu. org/auto- builds. This bug will be automatically closed when a new build containing the fix is uploaded to the current development release (Lucid Lynx).