What's the safest way to truncate the MySQL slow query log (under Linux primarily, but Windows would be handy to know) while MySQL is running?
By safe I mean:
Mustn't cause any permissions problems
Mustn't jump back to its original size next time its appended to
To truncate a file regardless if it is open by a running application do:
> /var/logs/your.log
at your shell prompt.
From here:
You can move the open file, then tell mysqld to flush logs, which will open a new log.
shell> cd mysql-data-directory
shell> mv mysql.log mysql.old
shell> mv mysql-slow.log mysql-slow.old
shell> mysqladmin flush-logs
If this is going to be a ongoing concern, you should take a look at logrotate as it can do this for you.
I did not know about echo > file.log, though it makes sense. I've always used cat /dev/null > file.log.
Also, it should be noted that it's very important not to delete a log file that is being written to because the program that has it open will continue to write to the file. Very difficult to figure out why all your hard drive space is gone!
Keep in mind that just deleting (rm) or moving (mv) the file will not free up space until mysql restart.
On ubuntu and MariaDB i used this to empty the disk space used from mysql-slow.log without restarting.
echo > /var/lib/mysql/mysql-slow.log
Related
I am not a MySQL expert.
I have a script that installs MySQL, starts mysqld, and then uses mysql to do some initialization.
Currently, in order to have this work, I enter into a loop that (apologize for the pseudocode mixing multiple languages):
mysqld_safe /* ... */ & /* ampersand to start in background so we can continue */
while(fileDoesNotExist("/tmp/mysql.sock")) {
sleepFor100ms();
}
mysql -u root /* and so forth */ initialize.sql
This seems to work (!) but has multiple problems:
polling smells funny,
I am not smart enough about MySQL to know whether looking at that hard-coded pathname /tmp/mysql.sock is smart at all.
And yet it's a lot easier than trying to (for example) consume and parse the stdout (or is it stderr?) of mysqld_safe to figure out whether the server has started.
My narrow question is whether there's a way to issue a blocking start of mysqld: can I issue any command that blocks until the database has started, and then exits (and detaches, maybe leaving a PID file), and has a companion stop command? (Or maybe allows me to read the PID file and issue my own SIGTERM?)
My broader question is, am I on the right track, or is there some totally different and easier (to be "easier" for me it would have to be lightweight; I'm not that interested in installing a bunch of tools like Puppet or DbMaintain/Liquibase or whatever) approach to solving the problem I articulated? That is, starting with a .gz file containing MySQL, install a userland MySQL and initialize a database?
Check out the init shell script for mysqld. They do polling, in a function called wait_for_pid().
That function checks for the existence of the pid file, and if it doesn't exist yet, sleeps for 1 whole second, then tries again. There's a timeout that defaults to 900 seconds, at which point it gives up waiting and concludes that it's not going to start (and outputs a totally unhelpful message "The server quit without updating PID file").
You don't have to guess where the pid file is. If you're starting mysqld_safe, you should tell it where it should create the pid file, using the --pid-file option.
One tricky part is that the pid file isn't created until mysqld initializes. This can take a while if it has to perform crash recovery using the InnoDB log files, and the log files are large. So it could happen that 900 seconds of timeout isn't long enough, and you get a spurious error, even though mysqld successfully starts a moment after the timeout.
You can also read the error log or the console output of mysqld. It should eventually output a line that says "ready for connections."
To read until you get this line, and then terminate the read, you could use:
tail -f | sed -e '/ready for connections/q'
You can use
mysqladmin -h localhost status
or use a pure bash solution like wait-for-it
./wait-for-it.sh --timeout 10 -h localhost -p 3306
I write a Web application using Yii framework and MySQL.
Now the boss wants "to store all changes in the database", in order to be able to restore older data if someone destroys some important information in the current version of the data.
What is to store all changes in the database is vague. I am not sure what exactly we should do.
How to fulfill this vague boss's requirement?
Can we do it with MySQL logs? What are pros and contras of using MySQL logs for this? Is it true that we need a programmer (me) to restore some (possibly not all) data from MySQL logs? Can MySQL (partial) data restoration be made simple?
Or should I hard work to manually (not with MySQL logs) store all old data in specific MySQL tables?
I guess what you are describing is an audit trail, which will be handy to go back and look at the history, but as for restoring, that will need to be manual.
Have a look at techniques for creating an audit trail.
You might want to try searching the extensions library for something like eactsasversioned that will archive edits made to records. I'm not sure if it saves deleted records, but it seems like it's close to what you want.
If you are looking for something you can easily restore from you probably need a backup script run on a very regular basis. I use a bash script(shown below) in cron to backup the databases I am worried about hourly. My databases are fairly small so this only takes a few seconds and could be increased to run every 15 minutes if you are super paranoid.
#!/bin/bash
dbName1="li_appointments"
dbName2="lidb_users"
dbName3="orangehrm_li"
fileName1=$dbName1"_`date +%Y.%m.%d-%H:%M:%S`.sql"
fileName2=$dbName2"_`date +%Y.%m.%d-%H:%M:%S`.sql"
fileName3=$dbName3"_`date +%Y.%m.%d-%H:%M:%S`.sql"
backupDir="/home/backups/mysql"
mysqldump -u backup_user --password='********************************' $dbName1 > $backupDir/$fileName1
mysqldump -u backup_user --password='********************************' $dbName2 > $backupDir/$fileName2
mysqldump -u backup_user --password='********************************' $dbName3 > $backupDir/$fileName3
bzip2 $backupDir/$fileName1
bzip2 $backupDir/$fileName2
bzip2 $backupDir/$fileName3
gpg -c --passphrase '********************************' $backupDir/$fileName1".bz2"
gpg -c --passphrase '********************************' $backupDir/$fileName2".bz2"
gpg -c --passphrase '********************************' $backupDir/$fileName3".bz2"
rm $backupDir/*.bz2
echo "Backups completed on `date +%D`" >> $backupDir/backuplog.log
If this is recommended ?
Can I ask some git command examples about how to track versions of mysql schema?
Should we use another repository other then the one we normally use on our application root ?
Should I use something called hook ?
Update:
1) We navigate onto our project root where .git database resides.
2) We create a sub folder called hooks.
3) We put something like this inside a file called db-commit:
#!/bin/sh
mysqldump -u DBUSER -pDBPASSWORD DATABASE --no-data=true> SQLVersionControl/vc.sql
git add SQLVersionControl/vc.sql
exit 0
Now we can:
4) git commit -m
This commit will include a mysql schema dump that has been run just before the commit.
The source of the above is here:
http://edmondscommerce.github.io/git/using-git-to-track-db-schema-changes-with-git-hook.html
If this is an acceptable way of doing it, can I please ask someone with patience to comment line by line and with as much detail as possible, what is happening here:
#!/bin/sh
mysqldump -u DBUSER -pDBPASSWORD DATABASE --no-data=true> SQLVersionControl/vc.sql
git add SQLVersionControl/vc.sql
exit 0
Thanks a lot.
Assuming you have a git repo already, do the following in a shell script or whatever:
#!/bin/bash -e
# -e means exit if any command fails
DBHOST=dbhost.yourdomain.com
DBUSER=dbuser
DBPASS=dbpass # do this in a more secure fashion
DBNAME=dbname
GITREPO=/path/to/git/repo
cd $GITREPO
mysqldump -h $DBHOST -u $DBUSER -p$DBPASS -d $DBNAME > $GITREPO/schema.sql # the -d flag means "no data"
git add schema.sql
git commit -m "$DBNAME schema version $(`date`)"
git push # assuming you have a remote to push to
Then start this script on a daily basis from a cron job or what have you.
EDIT: By placing a script in $gitdir/hooks/pre-commit (the name is important), the script will be executed before every commit. This way the state of the DB schema is captured for each commit, which makes sense. If you automatically run this sql script every time you commit, you will blow away your database, which does not make sense.
#!/bin/sh
This line specifies that it's a shell script.
mysqldump -u DBUSER -pDBPASSWORD DATABASE --no-data=true> SQLVersionControl/vc.sql
This is the same as in my answer above; taking the DDL only from the database and storing it in a file.
git add SQLVersionControl/vc.sql
This adds the SQL file to every commit made to your repository.
exit 0
This exits the script with success. This is possibly dangerous. If mysqldump or git add fails, you may blow away something you wanted to keep.
If you're just tracking the schema, put all of the CREATE statements into one .sql file, and add the file to git.
$> mkdir myschema && cd myschema
$> git init
$> echo "CREATE TABLE ..." > schema.sql
$> git add schema.sql
$> git commit -m "Initial import"
IMO the best approach is described here: http://viget.com/extend/backup-your-database-in-git. For your convenience I repeat the most important pieces here.
The trick is to use mysqldump --skip-extended-insert, which creates dumps that can be better tracked/diffed by git.
There are also some hints regarding the best repository configuration in order to reduce disk size. Copied from here:
core.compression = 9 : Flag for gzip to specify the compression level for blobs and packs. Level 1 is fast with larger file sizes, level 9 takes more time but results in better compression.
repack.usedeltabaseoffset = true : Defaults to false for compatibility reasons, but is supported with Git >=1.4.4.
pack.windowMemory = 100m : (Re)packing objects may consume lots of memory. To prevent all your resources go down the drain it's useful to put some limits on that. There is also pack.deltaCacheSize.
pack.window = 15 : Defaults to 10. With a higher value, Git tries harder to find similar blobs.
gc.auto = 1000 : Defaults to 6700. As indicated in the article it is recommended to run git gc every once in a while. Personally I run git gc --auto everyday, so only pack things when there's enough garbage. git gc --auto normally only triggers the packing mechanism when there are 6700 loose objects around. This flag lowers this amount.
gc.autopacklimit = 10: Defaults to 50. Every time you run git gc, a new pack is generated of the loose objects. Over time you get too many packs which waste space. It is a good idea to combine all packs once in a while into a single pack, so all objects can be combined and deltified. By default git gc does this when there are 50 packs around. But for this situation a lower number may be better.
Old versions can be pruned via:
git rebase --onto master~8 master~7
(copied from here)
The following includes a git pre-commit hook to capture mysql database/schema, given user='myuser', password='mypassword', database_name='dbase1'. Properly bubbles errors up to the git system (the exit 0's in other answers could be dangerous and may not handle error scenarios properly). Optionally, can add a database import to a post-checkout hook (when capturing all the data, not just schema), but take care given your database size. Details in bash-script comments below.
pre-commit hook:
#!/bin/bash
# exit upon error
set -e
# another way to set "exit upon error", for readability
set -o errexit
mysqldump -umyuser -pmypassword dbase1 --no-data=true > dbase1.sql
# Uncomment following line to dump all data with schema,
# useful when used in tandem for the post-checkout hook below.
# WARNING: can greatly expand your git repo when employing for
# large databases, so carefully evaluate before employing this method.
# mysqldump -umyuser -pmypassword dbase1 > dbase1.sql
git add dbase1.sql
(optional) post-checkout hook:
#!/bin/bash
# mysqldump (above) is presumably run without '--no-data=true' parameter.
set -e
mysql -umyuser -pmypassword dbase1 < dbase1.sql
Versions of apps, OS I'm running:
root#node1 Dec 12 22:35:14 /var/www# mysql --version
mysql Ver 14.14 Distrib 5.1.54, for debian-linux-gnu (x86_64) using readline 6.2
root#node1 Dec 12 22:35:19 /var/www# git --version
git version 1.7.4.1
root#node1 Dec 12 22:35:22 /var/www# lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 11.04
Release: 11.04
Codename: natty
root#node1 Dec 12 22:35:28 /var/www#
While I am not using Git, I have used source control for over 15 years. A best practice to adhere to when deciding where and how to store your src and accompanying resources in Source Control: If the DB Schema is used within the project then you should be versioning the schema and all other project resources in "that" project. If you develop a set of schemas or programming resources that you resuse in other projects then you should have a seperate repository for those reusable resources. That seperate Reusable resources project will be versioned on it's own and will track the versions of the actual reusable resources in that repository.
If you use a versioned resource out of the reusable repository in a different project then you have the following scenario, (just an example). Project XYZ version 1.0 is now using DB Schema_ABC version 4.0 In this case you will understand that you have used a specific version of a reusable resource and since it is versioned you will be able to track its use throughout your project. If you get a bug report on DBSchema_ABC, you will be able to fix the schema and re-version as well as understand where else DBSchem_ABC is used and where you may have to make some changes. From there you will also understand which projects contain wich versions of which reusable resources... You just have to understand how to track your resources.
Adopting this type of development Environment and Resource Management strategy is key to releasing usable software and managing a break/fix enhancement environment. Even if you're developing for your own edificcation on your own time, you should be using source control.. as you are..
As for Git, I would find a gui front end or a dev env integration if I can. Git is pretty big so I am sure it has plenty of front end support, maybe?
As brilliant as it sounds (the idea did occur to me as well), when I tried to implement it, I hit a wall. In theory, by using the --skip-extended-insert flag, despite initial dump would be big, the diffs between daily dumps should be minimal, hence the size increase over time of the repository could be assumed to be minimal as well, right? Wrong!
Git stores shapshots, not diffs, which means on each commit, it will take the entire dump file, not just the diff. Moreover, since the dump with --skip-extended-instert will use all field names on every single insert line, it will be huge compared to a dump done without --skip-extended-instert. This results in an explosion in size, the exact opposite what one would expect.
In my case, with a ~300MB sql dump, the repository went to gigabytes in days. So, what did I do? I first tried the same thing, only remove --skip-extended-instert, so that dumps will be smaller, and snapshots would be proportionally smaller as well. This approach held for a while, but in time it became unusable as well.
Still, the diff usage with --skip-extended-insert actually still seemed like a good idea, only, now I try to use subversion instead of git. I know, compared to git, svn is ancient history, yet it seems to work better, since it actually does use diffs instead of snapshots.
So in short, I believe best solution is doing the above, but with subversion instead of git.
(shameless plug)
The dbvc commandline tool allows you to manage your database schema updates in your repository.
It creates and uses a table _dbvc in the database which holds a list of the updates that are run. You can easily run the updates that haven't been apply to your database schema yet.
The tool uses git to determine the correct order of executing the updates.
DBVC usage
Show a list of commands
dbvc help
Show help on a specific command
dbvc help init
Initialise DBVC for an existing database.
dbvc init
Create a database dump. This is used to create the DB on a new environment.
mysqldump foobar > dev/schema.php
Create the DB using the schema.
dbvc create
Add an update file. These are used to update the DB on other environments.
echo 'ALTER TABLE `foo` ADD COLUMN `status` BOOL DEFAULT 1;' > dev/updates/add-status-to-foo.sql
Mark an update as already run.
dbvc mark add-status-to-foo
Show a list of updates that need to be run.
dbvc status
Show all updates with their status.
dbvc status --all
Update the database.
dbvc update
I have found the following options to be mandatory for a version control / git-compatible mysqldump.
mysqldump --skip-opt --skip-comments |sed -e 's/DEFINER[ ]*=[ ]*[^*]*\*/\*/'
(and maybe --no-data)
--skip-opt is very useful, it takes away all of --add-drop-table --add-locks --create-options --disable-keys --extended-insert --lock-tables --quick --set-charset. The DEFINER sed is necessary when the database contains triggers.
I'm trying make a copy of a MySQL database on another server. I stop the server, tar up the mysql directory, copy it to the other server and untar it. I set all the permissions to match to the working server, and copied the my.cnf file so everything is the same except for server specific changes.
However, when I try to startup the server, I get the following InnoDB error:
InnoDB: Operating system error number 13 in a file operation.
This error means mysql does not have the access rights to
the directory.
File name /var/lib/mysql/ibdata1
File operation call: 'open'.
The owner/group for all the files is mysql. I even tried changing permissions to a+rw. I can su to the mysql user and access the ibdata1 file using head.
SOLUTION:
The problem was selinux was enabled and preventing the new files from being accessed.
A silly question, but people forget: you said you checked that all files have the same permissions; still, even though it said so in the message, might you possibly have forgotten to check the permissions on the containing directory?
UPDATE: Two more suggestions:
You might try inserting --console and --log-warnings flags, for lots of debugging output, something like this (on my Mac):
/usr/libexec/mysqld --console --log-warnings --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/var/run/mysqld/mysqld.pid --skip-external-locking --socket=/var/lib/mysql/mysql.sock
If all else fails, you can probably try strace mysqld ... to see what exactly is it failing. The error will be somewhere at the bottom.
UPDATE2: Interesting indeed... I can't see what your OS is. I normally don't use /sbin/service, it's a bit mysterious for me; on a Mac, it's deprecated in favour of launchctl with config file in /System/Library/LaunchDaemons/mysqld.plist, and on most Linux boxes you have /etc/init.d/mysqld. So you could insert strace there.
Or (untested, but manpage suggests it's possible) you could try stracing the service call:
strace -ff -o straces /sbin/service start mysqld
This should produce files straces.pid, one of which should be mysqld's, and hopefully you'll find your error there.
This isn't a direct answer to your question, but I would recommend trying one of these programs for your backup / restore needs.
Percona Xtrabackup: https://launchpad.net/percona-xtrabackup
Mydumper: http://www.mydumper.org/
Both are great tools, are free and open source, and will help you avoid that problem entirely.
Hope that helps.
I'm trying to set up a MySQL database so that the tables are ran by the memory engine. I don't really care about loosing some data that gets populated but I would like to dump it daily (via mysqldump in a cronjob) and have the init-file set to this dump. However I can't seem to figure out how to get the mysqldump to be compatable with how the init-file wants the SQL statements to be formatted.
Am I just missing something completely obvious trying to set up a database this way?
MySQL dumps are exactly that -- dumps of the MySQL database contents as SQL. So, there isn't any way to read this directly as a database file.
What you can do, is modify your init script for MySQL to automatically load the last dump (via the command line) every time MySQL starts.
An even better solution would be to use a ramdisk to hold the entire contents of your database in memory, and then periodically copy this to a safe location as your backup.
Although, if you want to maintain the contents of your databases at all, you're better off just using one of the disk-based storage engines (InnoDB or MyISAM), and just giving your server a lot of RAM to use as a cache.
This solution is almost great, but it causes problems when string values in table data contain semicolons - all of them are replaced with newline char.
Here is how I implemented this:
mysqldump --comments=false --opt dbname Table1 Table2 > /var/lib/mysql/mem_tables_init.tmp1
#Format dump file - each statement into single line; semicolons in table data are preserved
grep -v -- ^-- /var/lib/mysql/mem_tables_init.tmp1 | sed ':a;N;$!ba;s/\n/THISISUNIQUESTRING/g' | sed -e 's/;THISISUNIQUESTRING/;\n/g' | sed -e 's/THISISUNIQUESTRING//g' > /var/lib/mysql/mem_tables_init.tmp2
#Add "USE database_name" instruction
cat /var/lib/mysql/mem_tables_init.tmp2 |sed -e 's/DROP\ TABLE/USE\ `dbname`;\nDROP\ TABLE/' > /var/lib/mysql/mem_tables_init.sql
#Cleanup
rm -f /var/lib/mysql/mem_tables_init.tmp1 /var/lib/mysql/mem_tables_init.tmp2
My understanding is that the --init-file is expecting each SQL statement on a single line and that there are no comments in the file.
You should be able to clear up the comments with:
mysqldump --comments=false
As for each SQL statement on one line, I'm not familiar with a mysqldump option to do that, but what you can do is a line of Perl to remove all of the newlines:
perl -pi -w -e 's/\n//g;' theDumpFilename
I don't know if --init-file will like it or not, but it's worth a shot.
The other thing you could do is launch mysql from a script that also loads in a regular mysqldump file. Not the solution you were looking for, but it might accomplish the effect you're after.
I stumbled onto this, so I'll tell you what I do. First, I have an ip->country db in a memory table. There is no reason to try to "save" it, its easily and regularly dropped and recreated, but it may be unpredictable how the php will act when its missing and its only scheduled to be updated weekly. Second, I have a bunch of other memory tables. There is no reason to save these, as they are even more volatile, with lifespans in minutes. They will be refreshed very quickly, but stale data is better than none at all. Also, if you are using any separate key caches, they may (in some cases) need to loaded first or you will be unable to load them. And finally, be sure to put a "use" statement in there if you're not dumpling complete databases, as there is no other interface (like mysql client) to open the database at start up.. So..
cat << EOF > /var/lib/mysql/initial_load.tmp
use fieldsave_db;
cache index fieldsave_db.search in search_cache;
EOF
mysqldump --comments=false -udrinkin -pbeer# fieldsave_db ip2c \
>> /var/lib/mysql/initial_load.tmp
mysqldump --comments=false -ufields -pavenue -B memtables \
>> /var/lib/mysql/initial_load.tmp
grep -v -- ^-- /var/lib/mysql/initial_load.tmp |tr -d '\012' \
|sed -e 's/;/;\n/g' > /var/lib/mysql/initial_load.sql
As always, YMMV, but it works for me.