Details
-
Type:
Bug
-
Status: Closed
-
Priority:
Major
-
Resolution: Not a Bug
-
Affects Version/s: 5.5.25-galera
-
Fix Version/s: None
-
Component/s: None
-
Labels:
-
Environment:Debian Squeeze amd64 on a qemu VM.
Description
I attached the complete terminal steps and the small tutorial I made. In /var/lib/mysql/kooka1h.err :
121005 15:18:22 mysqld_safe Starting mysqld daemon with databases from /var/lib/mysql
121005 15:18:22 [ERROR] An old style --language value with language specific part detected: /usr/share/mysql/english/
121005 15:18:22 [ERROR] Use --lc-messages-dir without language specific part instead.
121005 15:18:22 [ERROR] mysqld got signal 11 ;
This could be because you hit a bug. It is also possible that this binary
or one of the libraries it was linked against is corrupt, improperly built,
or misconfigured. This error can also be caused by malfunctioning hardware.
To report this bug, see http://kb.askmonty.org/en/reporting-bugs
We will try our best to scrape up some info that will hopefully help
diagnose the problem, but since we have already crashed,
something is definitely wrong and this may fail.
Server version: 5.5.25-MariaDB
key_buffer_size=0
read_buffer_size=131072
max_used_connections=0
max_threads=153
thread_count=0
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 335099 K bytes of memory
Hope that's ok; if not, decrease some variables in the equation.
Thread pointer: 0x0x0
Attempting backtrace. You can use the following information to find out
where mysqld died. If you see no messages after this, something went
terribly wrong...
stack_bottom = 0x0 thread_stack 0x30000
Gliffy Diagrams
Attachments
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Hi,
I installed a clean Debian Squeeze amd64 in a VM and executed the provided instruction, step by step (with the only tiny correction: it says "wsrep_provider=/usr/lib/galera", but this is a folder, the correct value would be "wsrep_provider=/usr/lib/galera/libgalera_smm.so". The server starts all right. And even with the wrong version of wsrep_provider, it doesn't crash, instead it complains in the log about the problem quite clearly and aborts.
However, some important information is missing here. The document says that you modified the my.cnf file, and both the lines you added and the error log confirm it. It means that you have remains of an unknown distribution/version – at the very least the config file, but since you didn't mention anything about uninstalling a previous version – quite possibly something else. Taking into account that you installed your mariadb-galera manually, you might have a weird mix of different mysql-related binaries and libraries on the machine.
If you still haven't resolved the problem, please clean up your system and try to install the cluster again. If the problem persists, please try to build the server with -DCMAKE_BUILD_TYPE=Debug, maybe it will give a clearer picture.
Please, can you show me the steps you made ? It will be very useful to compare it to my own steps.
Thanks
I meticulously followed your instruction from the RTF file, step by step, except for:
1) root@serv1: wget http://downloads.mariadb.org/f/mariadb-galera-5.5.25/kvm-tarbake-jaunty-x86/mariadb-5.5.25.tar.gz/from/http:/mirror3.layerjet.com/mariadb
Didn't work quite like that for me, so I had to download it in a different way; but I used the same mirror and the same source package.
2) Créer ces lignes:
deb http://ftp.debian.org/debian squeeze main contrib
deb http://security.debian.org squeeze/updates main contrib
deb-src http://ftp.debian.org/debian squeeze main contrib
deb-src http://security.debian.org squeeze/updates main contrib
I already had some of those in the file (the brand new one, from the fresh installation), so apt-get complained about duplicates and I had to remove the duplicates
3) root@serv1: nano /etc/mysql/my.cnf
ajouter ou modifier:
nano complained that the file didn't exist, so I had to create it first.
Then I added [mysqld] section and put there the lines that you provided.
4) At the point where you say "Démarrer le serveur puis:" I ran /etc/init.d/mysql
The server started all right.
That's about it.
As I said before, the main difference that I see is that you already had some MySQL/MariaDB installation on your machine, and I didn't.
That works very well except for the init script. It doesn't accept the start command. I have that line in my error log:
/usr/local/mysql/bin/mysqld: Too many arguments (first extra is 'start').
121122 18:26:07 [ERROR] Aborting
'/usr/local/mysql/bin/mysqld' alone works but '/usr/local/mysql/bin/mysqld start' not.
The problem is I call the init script like that:
/etc/init.d/mysql start
Thanks.
I made my own init script which works but I have another problem:
121122 19:21:44 [Note] WSREP: Running: 'wsrep_sst_rsync_wan 'joiner' '10.111.200.9' '' '/usr/local/mysql/data/' '/etc/mysql/my.cnf' '1329''
sh: wsrep_sst_rsync_wan: not found
121122 19:21:44 [ERROR] WSREP: Failed to read 'ready <addr>' from: wsrep_sst_rsync_wan 'joiner' '10.111.200.9' '' '/usr/local/mysql/data/' '/etc/mysql/my.cnf' '1329'
Read: '(null)'
when I try to connect a server to the first server.
Can you help me ?
Thanks again.
Regarding the script, if I remember correctly (I don't have that installation currently) your algorithm causes /etc/init.d/mysql to be a link to mysqld_safe, which of course doesn't have 'start' command.
For the error above, it's just what it says: there is no wsrep_sst_rsync_wan script in the current package. Try to use wsrep_sst_method=rsync instead of wsrep_sst_method=rsync_wan
It works !!!! Thanks a lot.
I will close the report then, please open a new one if you observe further problems
There is really not much here to go with, all we can do is to check installation of the binary package on Debian Squeeze.