maria-developers team mailing list archive

[Roberto Spadim <roberto@xxxxxxxxxxxxx>]

data partially recovered, procedure:


aria_read_log -r --datadir=/home/mysql/data/local/ --require-control-file
-e rastreabilidade --transaction-log --start-from-lsn=67
..... ok
aria_read_log -r --datadir=/home/mysql/data/local/ --require-control-file
-e rastreabilidade --transaction-log --start-from-lsn=68
..... ok

aria_chk -r --datadir=/home/mysql/data/local/ --require-control-file -e
--transaction-log <table names>




mysql start....
segment fault (same problem: 2016-12-22  3:41:13 139878188566400 [Note]
mysqld: Aria engine: starting recovery,  recovered pages: 0% 3082852397%
0%161222  3:41:13 [ERROR] mysqld got signal 11 ;)

rm aria_log.00000067

mysql start....
segment fault (same problem: 2016-12-22  3:41:13 139878188566400 [Note]
mysqld: Aria engine: starting recovery,  recovered pages: 0% 3082852397%
0%161222  3:41:13 [ERROR] mysqld got signal 11 ;)

rm aria_log.00000068

mysql start....
OK, but some data was not ok (maybe data after checkpoint)


----------------
i didn't understood why mysqld force a "aria_read_log" with
--start-from-checkpoint and crashed, if you check this :
recovered pages: 0% 3082852397% 0%Segmentation fault

it have >100% values?!!?!? (3082852397%)



!!should be nice a option at MY.CNF to FORCE aria_read_log with
'START_FROM_LSN' instead of 'start_from_checkpoint'!!

should i get all this emails and send to JIRA issue tracker?