Hi Thomas,
The topmost reason for this kind of failure is usually an existing
my.cnf file which contains option(s) not supported by the current
version of the server, e.g. deprecated and removed. You mentioned that
you purged everything mysql-ish, did it include all instances of my.cnf
you could find? They could be hiding in /etc/my.cnf or /usr/etc/my.cnf, etc.
Further, if you don't have syslog, your error output currently just goes
nowhere; at least I tried to stop rsyslog, and that's what happens.
Please try to modify /etc/mysql/conf.d/mysqld_safe_syslog.cnf -- it most
likely has 'syslog' option, just change it to skip-syslog. It should
enable default error logging to /var/lib/mysql/<machine name>.err. Then,
try to run /etc/init.d/mysql start again and see if the error log is
updated.
If it still isn't, I'd recommend to check permissions on /var/lib/mysql.
Regards,
Elena
On 2/4/2013 10:38 PM, Thomas Hackert wrote:
Hello Kristian, *,
On Mon, Feb 04, 2013 at 10:27:59AM +0100, Kristian Nielsen wrote:
Thomas Hackert <[email protected]> writes:
It seems, that it is a problem to try to stop a non-installed server
before starting it, or am I wrong ;?
It does not look like the problem is stopping the server. The error is about a
failure to *start* the server.
O.K.
I had tested it a little bit further. My findings so far:
1. I started the purging of mariadb and its dependencies again.
2. Started to install mariadb-server again. Still the same error message
... :(
3. Purged it again, removed all founded mysql files and directories, and
tried to install mariadb-server again, but to no avail ... :(
Ok, I was going to suggest moving away (or deleting if you don't need it)
anything in /etc/mysql/ and /var/lib/mysql/ and /var/log/mysql/. But seems you
already tried this, and it did not work...
Unfortunately not ... :(
/If/ there would be any ... :( Would it be of any help, if I install a
logging daemon (not sure, why I have no syslogd and/or alternatives of
it installed ... :( I was pretty sure, that some progs need to log every
Hm, I wonder if this is related. The server logs messages to syslog for .deb
packages. I am not sure of the exact mechanism though. I wonder if the server
tries to log to syslog, this gives an error because of no syslog present, and
this is what causes the server start to fail?
Should syslog not a dependency to install mariadb-server then?
There is not much I can say without a better idea of why the server fails to
start. One thing you could try is to try and start the server manually after
the install fails:
sudo /etc/init.d/mysql start
Done this before (but had apparently forgotten to mention it, sorry ...
:( ). Leads to
<quote>
/etc/init.d/mysql start
[....] Starting MariaDB database server: mysqld . . . . . . . . . . . .
. . . . . . . . [FAIL . . . . . . . failed!
</quote>
... :(
You can try with strace -f to spot why it fails, or you can edit
/etc/init.d/mysql to make it more verbose in an attempt to spot what causes
the problem. Stuff like that.
Well, I have "strace"d it to a file, but am not sure, what it is telling
me ... :( The file is 7.2 MB ("bzip"ped only 340 KB) ... :( Is it
possible to send it to someone, who is able to "read" strace? I am not,
sorry ... :(
Or is there a way for me to reproduce the problem myself?
I am not sure. I have only added mariadb's repository to my
/etc/apt/sources.list, run "apt-get update" and tried to install
mariadb-server, when I stumbled upon this problem ... ;)
Thanks for your answers and have a nice evening
Thomas.
_______________________________________________
Mailing list: https://launchpad.net/~maria-developers
Post to : [email protected]
Unsubscribe : https://launchpad.net/~maria-developers
More help : https://help.launchpad.net/ListHelp
