RT and mysql replication

Hi.

I’d googled about this quesitons, but I’d find no useful information.
You can simply tell me RTFM, but please link the FM I must R… :wink:

Is there an RT support for mysql replication, a single master/many slave
databases simple clustering method described in this page?

http://dev.mysql.com/doc/refman/5.0/en/replication.html

Luca Villani Mobile Team, Dada S.p.A.
Tel: +39 055 2267220 Mob: +39 335 8753086
ICQ: 76272621 Skype: luca.villani
GPG key fingerprint: 7FC9 E2FE 0BEE 9DF8 1719 8761 1B79 82CC F0B5 B7CF

Hi.

I’d googled about this quesitons, but I’d find no useful information.
You can simply tell me RTFM, but please link the FM I must R… :wink:

Is there an RT support for mysql replication, a single master/many slave
databases simple clustering method described in this page?

http://dev.mysql.com/doc/refman/5.0/en/replication.html

I don’t believe this is an RT question. Mysql supports replication,
for sure at 4.x, and supports clusters at 5.x. Do the needful
at the mysql level, RT doesn’t care. Just make sure to use a version
of mysql supported by RT. (We happen to use replication with 4.x,
but there is nothing in RT that would depend on this.)

bobg

Alle 15:39, venerdì 4 maggio 2007, Bob Goldstein ha scritto:

I don’t believe this is an RT question. Mysql supports replication,
for sure at 4.x, and supports clusters at 5.x.

As far as in a replication cluster scenario applications must perform insert
and updates only onto the master db server and can perform select onto an
arbitrary slave server of the cluster, this will be an RT question: code must
be patched to follow this behaviour.

Do the needful at the mysql level, RT doesn’t care.

RT doesn’t care because it use only one cluster server: the master server.

We’re using RT in an heavy duty environment (about 1500/1600 tickets per day),
with about 90 different CRM operators, and using a single mysql server is not
a good deal: if RT can split searches onto some slaves db servers (4/5 slave
server) it can improve all operativeness.

Luca Villani Mobile Team, Dada S.p.A.
Tel: +39 055 2267220 Mob: +39 335 8753086
ICQ: 76272621 Skype: luca.villani
GPG key fingerprint: 7FC9 E2FE 0BEE 9DF8 1719 8761 1B79 82CC F0B5 B7CF

I know that this does not help your situation, but pgpool will allow
you to perform select queries to a slave postgresql DB and updates to
the master DB. There may be a similar proxy available for MySQL that
would allow you to scale your DB load better.

KenOn Fri, May 04, 2007 at 05:10:56PM +0200, Ham MI-ID, Torsten Brumm wrote:

Hi,
Just to pick up this discussion, this palnned scenario would be very, very clever, should be added to the rt4 wish list.

Torsten.

Btw: on on mysql cluster won’t help if you have on on very big environment, the replica architecture would be the best way i think.

-----Original Message-----
From: rt-users-bounces@lists.bestpractical.com rt-users-bounces@lists.bestpractical.com
To: rt-users@lists.bestpractical.com rt-users@lists.bestpractical.com
Sent: Fri May 04 16:27:05 2007
Subject: Re: [rt-users] RT and mysql replication

Alle 15:39, venerd?? 4 maggio 2007, Bob Goldstein ha scritto:

I don’t believe this is an RT question. Mysql supports replication,
for sure at 4.x, and supports clusters at 5.x.

As far as in a replication cluster scenario applications must perform insert
and updates only onto the master db server and can perform select onto an
arbitrary slave server of the cluster, this will be an RT question: code must
be patched to follow this behaviour.

Do the needful at the mysql level, RT doesn’t care.

RT doesn’t care because it use only one cluster server: the master server.

We’re using RT in an heavy duty environment (about 1500/1600 tickets per day),
with about 90 different CRM operators, and using a single mysql server is not
a good deal: if RT can split searches onto some slaves db servers (4/5 slave
server) it can improve all operativeness.


Luca Villani Mobile Team, Dada S.p.A.
Tel: +39 055 2267220 Mob: +39 335 8753086
ICQ: 76272621 Skype: luca.villani
GPG key fingerprint: 7FC9 E2FE 0BEE 9DF8 1719 8761 1B79 82CC F0B5 B7CF


The rt-users Archives

Community help: http://wiki.bestpractical.com
Commercial support: sales@bestpractical.com

Discover RT’s hidden secrets with RT Essentials from O’Reilly Media.
Buy a copy at http://rtbook.bestpractical.com


The rt-users Archives

Community help: http://wiki.bestpractical.com
Commercial support: sales@bestpractical.com

Discover RT’s hidden secrets with RT Essentials from O’Reilly Media.
Buy a copy at http://rtbook.bestpractical.com

Alle 17:30, venerdì 4 maggio 2007, Kenneth Marshall ha scritto:

I know that this does not help your situation, but pgpool will allow
you to perform select queries to a slave postgresql DB and updates to
the master DB. There may be a similar proxy available for MySQL that
would allow you to scale your DB load better.

Generally speaking, we prefer to put no wrapper/proxy on our architecture, due
to high load of our systems.

Do you have a link for a mysql proxy acting like this?

Luca Villani Mobile Team, Dada S.p.A.
Tel: +39 055 2267220 Mob: +39 335 8753086
ICQ: 76272621 Skype: luca.villani
GPG key fingerprint: 7FC9 E2FE 0BEE 9DF8 1719 8761 1B79 82CC F0B5 B7CF

Alle 17:30, venerd? 4 maggio 2007, Kenneth Marshall ha scritto:

I know that this does not help your situation, but pgpool will allow
you to perform select queries to a slave postgresql DB and updates to
the master DB. There may be a similar proxy available for MySQL that
would allow you to scale your DB load better.

Generally speaking, we prefer to put no wrapper/proxy on our architecture, due
to high load of our systems.

Do you have a link for a mysql proxy acting like this?

Since RT uses DBI to access the underlying database, you should be able
to use DBIx::DBCluster to send your writes to the master DB and read from
the slave(s).

Ken

Alle 18:26, venerdì 4 maggio 2007, Kenneth Marshall ha scritto:

Since RT uses DBI to access the underlying database, you should be able
to use DBIx::DBCluster to send your writes to the master DB and read from
the slave(s).

Do you know something on the net using DBIx::DBCluster for an RT patch?

Or are you suggest me to patch the code?

Luca Villani Mobile Team, Dada S.p.A.
Tel: +39 055 2267220 Mob: +39 335 8753086
ICQ: 76272621 Skype: luca.villani
GPG key fingerprint: 7FC9 E2FE 0BEE 9DF8 1719 8761 1B79 82CC F0B5 B7CF