Comment 3 for bug 938594

Revision history for this message
Matt (whereswardy) wrote :

The difference is that with 1.6.5 during the prepare step it would do a huge log recovery step that would take about 30 minutes, that looked like this:

120223 17:13:31 InnoDB: Database was not shut down normally!
InnoDB: Starting crash recovery.
InnoDB: Reading tablespace information from the .ibd files...
InnoDB: Doing recovery: scanned up to log sequence number 63858176 (1 %)
...
InnoDB: Doing recovery: scanned up to log sequence number 68738858 (99 %)
120223 17:13:32 InnoDB: Starting an apply batch of log records to the database...
InnoDB: Progress in percents: 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
InnoDB: Apply batch completed

It would be the "Progress in percents" step that would take about 99% of the prepare time before, and I'd always have to give it a lot of memory (typically 16GB) in order to get it to run at that speed (any less and it would take even longer).

This branch doesn't do that log recovery bit at all, but I still end up with a valid backup at the end.