Re: what is the best way to do unattended recovery of myisam tables?

[Michael Widenius]

>>>>> "Bill" == Bill Marrs <bill@stripped> writes:

>> The command is, as I said: cd your_database; myisamchk -fe *.MYI

Bill> -f will restart with -r if it finds there are any errors in the table.  My 
Bill> desire for this recovery to work the first time without a need for someone 
Bill> to re-run it (the --help for -o says "Slower than '-r' but can handle a 
Bill> couple of cases where '-r' reports that it can't fix the data file"), made 
Bill> me think that using -ef might not always fix everything.

Note that it's VERY unlikely that -r shouldn't be able to fix the
error.  The main difference between -r and -ro is that -r needs
disk space to sort indexes and -ro doesn't need this.  -r is also MUCH
faster than -ro.

Bill> Also, why isn't the functionality of -A (auto-increment repair) 
Bill> incorporated into -r and -o?

-A will set the auto_increment value to max(auto_incrmeent_value) +1

If you have deleted rows, this may not be the value you want and
because of this the -A is optional.

-r will however correct auto_increment values that are lower than
  max(auto_increment_value) +1

Regards,
Monty