How do you fix a MySQL "Incorrect key file" error when you can't repair the table?

I'm trying to run a rat开发者_高级运维her large query that is supposed to run nightly to populate a table. I'm getting an error saying Incorrect key file for table '/var/tmp/#sql_201e_0.MYI'; try to repair it but the storage engine I'm using (whatever the default is, I guess?) doesn't support repairing tables.

how do I fix this so I can run the query?

---

You must change the location of MySQL's temporary folder which is '/tmp' in most cases to a location with a bigger disk space. Change it in MySQL's config file.

Basically your server is running out of disk space where /tmp is located.

---

You'll need to run this command from the MySQL prompt:

REPAIR TABLE tbl_name USE_FRM;

From MySQL's documentation on the Repair command:

The USE_FRM option is available for use if the .MYI index file is missing or if its header is corrupted. This option tells MySQL not to trust the information in the .MYI file header and to re-create it using information from the .frm file. This kind of repair cannot be done with myisamchk.

---

Your query is generating a result set so large that it needs to build a temporary table either to hold some of the results or some intermediate product used in generating the result.

The temporary table is being generated in /var/tmp. This temporary table would appear to have been corrupted. Perhaps the device the temporary table was being built on ran out of space. However, usually this would normally result in an "out of space" error. Perhaps something else running on your machine has clobbered the temporary table.

Try reworking your query to use less space, or try reconfiguring your database so that a larger or safer partition is used for temporary tables.

MySQL Manual - B.5.4.4. Where MySQL Stores Temporary Files

---

The storage engine (MyISAM) DOES support repair table. You should be able to repair it.

If the repair fails then it's a sign that the table is very corrupted, you have no choice but to restore it from backups.

If you have other systems (e.g. non-production with same software versions and schema) with an identical table then you might be able to fix it with some hackery (copying the frm an MYI files, followed by a repair).

In essence, the trick is to avoid getting broken tables in the first place. This means always shutting your db down cleanly, never having it crash and never having hardware or power problems. In practice this isn't very likely, so if durability matters you may want to consider a more crash-safe storage engine.

---

Simple "REPAIR the table" from PHPMYADMIN solved this problem for me.

1. go to phpmyadmin
2. open problematic table
3. go to Operations tab (in my version of PMA)
4. at the bottom you will find "Repair table" link

---

In my case, there was a disc space issue. I deleted some unwanted war files from my server and it worked after that.

---

REPAIR TABLE tbl_name USE_FRM;

Command only run when MySQL 'Storage Engine' type should be 'MyISAM'

Hope this helps

---

this issue is because of low storage space availability of a particular drive(c:\ or d:\ etc.,), release some memory then it will work.

Thanks Saikumar.P

---

This happenes might be because you ran out of disk storage and the mysql files and starting files got corrupted

The solution to be tried as below

First we will move the tmp file to somewhere with larger space

Step 1: Copy your existing /etc/my.cnf file to make a backup

cp /etc/my.cnf{,.back-`date +%Y%m%d`}

Step 2: Create your new directory, and set the correct permissions

mkdir /home/mysqltmpdir
chmod 1777 /home/mysqltmpdir

Step 3: Open your /etc/my.cnf file

nano /etc/my.cnf

Step 4: Add below line under the [mysqld] section and save the file

tmpdir=/home/mysqltmpdir

Secondly you need to remove or error files and logs from the /var/lib/mysql/ib_* that means to remove anything that starts by "ib"

rm /var/lib/mysql/ibdata1 and rm /var/lib/mysql/ibda.... and so on

Thirdly you will need to make sure that there is a pid file available to have the database to write in

Step 1 you need to edit /etc/my.cnf

pid-file= /var/run/mysqld/mysqld.pid 

Step 2 create the directory with the file to point to

 mkdir /var/run/mysqld
 touch /var/run/mysqld/mysqld.pid
 chown -R mysql:mysql /var/run/mysqld

Last step restart mysql server

/etc/init.d/mysql restart

---

I just resolved a similar issue “Incorrect key file:\bonga_process\alarms.MYI’:try to repair it”

How I resolved it:

1. I started the server configuration application, ServerConfiguration.exe
2. Maitenance tab
3. Click on Repair tables command button
4. Prompts to delete files to free disk space appear with Manage Storage command button
5. Click on Manage Storage command, this will run a take you to some files you need to delete to free more disk space. Delete any useless files.
6. Restart the system auditor application

---

Apply proper charset and collation to database, table and columns/fields.

I creates database and table structure using sql queries from one server to another. it creates database structure as follows:

1. database with charset of "utf8", collation of "utf8_general_ci"
2. tables with charset of "utf8" and collation of "utf8_bin".
3. table columns / fields have charset "utf8" and collation of "utf8_bin".

I change collation of table and column to utf8_general_ci, and it resolves the error.

---

Change to MyISAM engine and run this command

REPAIR TABLE tbl_name USE_FRM;