Further to the email below,I performed the upgrade last night and we are now running mysql 4.0.16.

the last on the syncronization hasnt reduced any, so if anyone knows how else I can keep these two databases in sync, I would be very interested in knowing how!

now that we are running the 4.0 series, I would hope our options for this will be increased!

thanks in advance for any help you can provide...

-------------------------------------------
Many thanks for your reply!

I am currently syncronizing two MySQL servers (version 3.23.49) on a very high traffic website.

There are, at peak times upwards of 600 updates a second (and many many more selects)

During these times the slave database will fall out of sync, sometimes by several thousand seconds (im aware that this calculation is the 'time now - timestamp of last update from the master)



What is the status of the slave thread in show processlist?

If it is 'Locked' then the slave thread may be starving because of too
many (long) selects on the slave.
The status of the slave thread is 'Locked' for quite some time.  Just now
I have checked and a single query (normally very quick) was in Locked
status for 11 seconds.

If your my.cnf contains low-priority-updates and your slaves are serving
selects continuously, then the updates won't come through. This setting
is mentioned in the manual.

If you remove low-priority-updates, the updates will be processed sooner
but you'll get in trouble if you have long running selects. These will
delay all following selects on the slave when an update is waiting for
the long query to end.
Hmm, ok.  I dont have low-priority-updates in the config file (i did see
that in the manual)

Unfortunately I am in a position where we can not have lag on the
databases (people pay for live data, if their select gets run on the
slave and the results are delayed due to slave lag, they tend to be
pissed off) so is there anything I else I can do to reduce this while
still keeping the redundnacy of master/slave?  (we cant run the website
just on one DB, it falls over with 'can not create thread errors' when
we run out of memory :()

I am splitting up the database at the moment, and moving some of the
biggest tables on to a new server, and this week I will be upgrading
both these servers to mysql 4.0.16 which i understand will potentially
give us quite a performance increase.

Do you have any other suggestions that you may be able to offer?

Ive copied the my.cnf files from the master and the slave at the bottom
of this email if that may offer any clues to what else we can do.

many thanks in advance for any assistance you can give.
--
Jonathan Tullett


my.cnf on the MASTER:


# Here follows entries for some specific programs
[safe_mysqld]
err-log         = /var/log/mysql/mysql.err

# The MySQL server
[mysqld]
user            = mysql
pid-file        = /var/run/mysqld/mysqld.pid
socket          = /var/run/mysqld/mysqld.sock
port            = 3306
skip-locking
basedir         = /usr
datadir         = /var/lib/mysql
tmpdir          = /tmp
language        = /usr/share/mysql/english
set-variable    = key_buffer=384M
set-variable    = max_allowed_packet=1M
set-variable    = table_cache=512
set-variable    = sort_buffer=2M
set-variable    = record_buffer=2M
set-variable    = thread_cache=8
set-variable    = max_connections=400
set-variable    = max_user_connections=360
# Try number of CPU's*2 for thread_concurrency
set-variable    = thread_concurrency=8
set-variable    = myisam_sort_buffer_size=64M
#log            = /var/log/mysql/mysql.log
server-id       = 1
master-host     = 10.0.2.1
master-user     = replicate
master-password = replicate
master-connect-retry = 60
replicate-ignore-db = mysql
skip-slave-start

log-bin         = /mysql-log/db1-new-bin
log-slow        = /mysql-log/db1-new-slow.log



my.cnf on the SLAVE:

# Here follows entries for some specific programs
[safe_mysqld]
err-log         = /var/log/mysql/mysql.err

# The MySQL server
[mysqld]
user            = mysql
pid-file        = /var/run/mysqld/mysqld.pid
socket          = /var/run/mysqld/mysqld.sock
port            = 3306
skip-locking
basedir         = /usr
datadir         = /var/lib/mysql
tmpdir          = /tmp
language        = /usr/share/mysql/english
set-variable    = key_buffer=384M
set-variable    = max_allowed_packet=1M
set-variable    = table_cache=512
set-variable    = sort_buffer=2M
set-variable    = record_buffer=2M
set-variable    = thread_cache=8
set-variable    = max_connections=400
set-variable    = max_user_connections=360
# Try number of CPU's*2 for thread_concurrency
set-variable    = thread_concurrency=8
set-variable    = myisam_sort_buffer_size=64M
#log            = /var/log/mysql/mysql.log
server-id       = 2
master-host     = 10.0.2.2
master-user     = replicate
master-password = replicate
master-connect-retry = 60
replicate-ignore-db = mysql
skip-slave-start

log-bin         = /mysql-log/db2-bin
log-slow        = /mysql-log/db2-slow.log

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]


________________________________________________________________________ This e-mail has been scanned for all viruses by Star Internet.

Should you have any problems regarding this service, please contact
[EMAIL PROTECTED]



--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:    http://lists.mysql.com/[EMAIL PROTECTED]



Reply via email to