On Tue, Sep 14, 2010 at 2:04 PM, Uwe Brauer oub@mat.ucm.es wrote:
On Tue, 14 Sep 2010 00:38:17 +0200, Platonides Platonides@gmail.com wrote:
> Uwe Brauer wrote: >> most of the files are very small in size with the >> exception of searchindex.MYD and searchindex.MYI. I >> have no idea whether this is "normal".
>> Well I run strings and grep over these files but the >> only files with useful information are searchindex.MYD >> and searchindey.MYI >>
> Your tables will be in InnoDB, with the table files > just holding pointers to ibdata1 file. searchindex > table is in MyIsam and thus the content is really on > those files.
To be honstest I don't understand completely what you ay, since I don't know exactly what InnoDB is although I have heard the word.
I just found a mysql dump file from another wiki. I could succesfully use that dump and later conncect wikimedia to that db. There is however a difference this db has more .MYD and MYI files for example wp_users.MYD wp_users.MYI
Etc.
Does this mean that the database from the crashed hard disk is may be corrupt, since it has only 1 MYD and 1 MYI file????
Is there any way to extract the ASCII text (anything more sophisticated than "strings") from searchindex.MYD and searchindey.MYI
>> > Look at the Mediawiki logfiles, see if they're >> > indicating any database problems. If they are... >> >> > Log in to the database with the mysql command line >> > command and run the command "show tables;" sorry my ignorance
You mean
mysql -p wikidb
mysql > show tables
Well it does not return anything also mediawiki connects correctly to that database.
I am puzzled
The "show tables;" command within MySQL has to have the semicolon at the end, or it doesn't work.
What are you seeing in the logs that convinces you the connection from MediaWiki is working properly?