[MySQL 5.0.41] - Released

modem

Verified User
Joined
Apr 7, 2004
Messages
396
I just saw this earlier and have updated the RHEL3 RPM's for my CentOS 3 server.
 
Did you just download and run them from www.mysql..com? Because the CentOS RPM's aren't available, and from the look of it, it was released May 1, which leads me to believe it probably won't be? Although 5.0.37 was. hmmm

Thanks for the info.
 
Yes, I downloaded the statically linked RPM's and used webmin to install them and to ignore dependencies. I first removed 5.0.37 completely and then used the statically linked ones and viola... works great!
 
Hi,

I used

Code:
 mkdir mysql
 cd mysql
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-client-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-devel-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-server-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-shared-5.0.41-0.i386.rpm
 rpm -Uvh MySQL-*-5.0.41-0.i386.rpm
 /etc/rc.d/init.d/mysqld restart
cd ..
rm -rf mysql

and then recompiled php to use the new mysql client libraries.

Grant
 
I wanted to send along a follow up note. After upgrading to the latest MySQL, I recompiled PHP and all of the associated files with it by doing a ./build php_ap2 y command and then following that up by building php perl, frontpage, and zend. I restarted httpd and that part worked fine.

However, immediately afterwards I noticed that email wasn't coming in. I didn't think anything of this and ignored it for about 2 hours. I started getting a few phone calls from clients about email not coming through and then I re-examined the issue and backtracked. I checked and email WAS being delievered into the mail files on the server. As I could log into directadmin and watch the size of a person's pop/imap account grow.

Then I began trying to login via webmail and roundcube, but each of those indicated that there was *no* email and that the inboxes were empty. I knew otherwise because I could see the virtual pop account sizes grow. At this point I stopped and restarted both exim and dovecot, but neither of those worked.

Finally at last moment thought, I decided to recompile and install dovecot because I was running RC31 rather than the final release. So I grabbed the final release, compiled it using the proper commands as given by DirectAdmin. Finally after doing that I restarted exim and dovecot, and sure enough it worked! I was able to immediately download emails from the pop accounts AND send emails.

The problem that existed was that even a client like outlook would time out when looking for the server, and would not be able to send or receive email. Here is where it got really weird. If I logged onto the server (before recompiling dovecot) and created a new pop3 account, I could send and receive email just fine. When I recompiled/installed dovecot... the problem was solved.

I have notified John about this to see what might be done or reported to the dev teams of both products.

Brad
 
Hello,

That sounds like the /etc/exim.conf didn't get the dovecot patch correctly.
That would cause all emails to end up going to the old mbox location in /var/spool/virtual/domain.com/user instead of the dovecot/Maildir spot of /home/username/imap/domain.com/user/Maildir/new.
If email continue to be delivered into the mbox files, and dovecot is running, you won't see it. By re-running the install, the mbox files get converted yet again, all emails would have been copied into the dovecot location, thus you could read them.

So at this point, you'll need to confirm that the exim.conf did actually get patched..as if it failed the first time, it can still have failed the 2nd time. The convertor could give you the false impression that everything is fine.. but email may still end up in the mbox location.

1) Test an email to yourself to see if it arrives and is visible using dovecot. If so, then it's all fine.

2) if not, then run this guide to both update your exim.conf/exim.pl to the latest version, and be sure to run the patch to setup the exim.conf for dovecot.

John
 
John,

Actually I was running Dovecot 1.0 RC31 for several weeks just fine up until yesterday morning. It was sending and delivering new emails very normally and I can validate that it was using Dovecot. However, it was after I recompiled and upgraded to PHP 5.2.2 and recompiled Zend, frontpage, that email functionality ceased. When email functionality ceased, I had not modified anything regarding to Dovecot or Exim and didn't even realize emails weren't functioning till a short time later. As far as I know, compiling PHP should *not* affect emails.

Anyway after I recompiled / upgraded to Dovecot 1.0 final, did the email functionality become restored.

Bradley
 
I have just upgraded to MySQL 5.0.41 too and am having problems with exim, but am not using dovecot.
Exim starts up ok and appears to be working however no mail is actually being stored.
Does anyone have any idea how to fix this? I'm using CentOS 4 and Exim 4.60 and have tried updating the config file (http://help.directadmin.com/item.php?id=51) to no avail
 
Sorry! it was the firewall I installed. Thats the problem when you install lots of things at the same time!
 
Hi,

I used

Code:
 mkdir mysql
 cd mysql
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-client-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-devel-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-server-5.0.41-0.i386.rpm
 wget http://mirrors.dedipower.com/www.mysql.com/Downloads/MySQL-5.0/MySQL-shared-5.0.41-0.i386.rpm
 rpm -Uvh MySQL-*-5.0.41-0.i386.rpm
 /etc/rc.d/init.d/mysqld restart
cd ..
rm -rf mysql

and then recompiled php to use the new mysql client libraries.

Grant

I tried upgrading, but I got this error message:

Code:
ERROR: 1136  Column count doesn't match value count at row 1
070610  0:54:22 [ERROR] Aborting

070610  0:54:22 [Note] /usr/sbin/mysqld: Shutdown complete

Installation of system tables failed!

Any idea's ?
 
I fixed this by running the following commands:
Code:
mysqlcheck --check-upgrade --all-databases --auto-repair
mysql_fix_privilege_tables
 
mysqlcheck: Got error: 1045: Access denied for user 'root'@'localhost' (using password: NO) when trying to connect

Server down...php doesnt make!
 
You'll need to add the '-p' option to provide a password.
The root password is located in /usr/local/directadmin/scripts/setup.txt
 
Back
Top