[Vtigercrm-developers] Anyone set up HA or replication?

Stefan Warnat ich at stefanwarnat.de
Mon May 15 10:51:07 GMT 2017


Hey,

NDB clusters are a lot slower, then standalone mysql. (I Checked during
MySQL ~5.3, maybe it is better now)

To replicate the Database is not really a problem.
But you need to be sure that all tables will be within the InnoDB format (I
think, there was some problems)

With new GTID based replication this is also "easier" to setup/maintain.

The following description looks good, if I remember correctly:
http://www.fromdual.ch/sites/default/files/How_to_implement_GTID_replication.pdf

Setup HAProxy with two backend MySQL - Server with a "MySQL compatible"
configuration. That is the hardest part, because if I remember correctly
the default config is not compatible and will blocked by mysql server.
HAProxy automatically switch Database to other server if there is a problem
only with database. (If you wants to have full server replication, you
maybe don't need HAProxy)

Real problem is the step back to Live system, because for this the old
"Live" must be configured as "Slave" to get all records, before you could
switch back.

For me only the files was a problem to replication correctly and I don't
wanted to spend more time in this topic. I used replication only for
fallback option of database during updates.
In the meantime filesystems, which automatically replicate are a lot
better. But now I don't need it anymore.

At all I must say, currently I'm not using this setup anymore, because it
don't runs smoothly during updates. And on some point between Debian 8 and
Debian 9 I lost the slave ;)

Regards,

Stefan


----
*Freelancer at Webdevelopment*

Web: http://www.stefanwarnat.de
Xing: https://www.xing.com/profile/Stefan_Warnat2
LinkedIn: http://www.linkedin.com/pub/stefan-warnat/6/827/820
eMail: kontakt at stefanwarnat.de


On Mon, May 15, 2017 at 9:56 AM, Conrado Maggi <comaggi at gmail.com> wrote:

> Hi,
>
> You could also check amazon RDS that offer replication out of the box.
> https://aws.amazon.com/rds/faqs/#replication
>
> Regards,
> Conrado
>
> On Fri, May 12, 2017 at 3:57 PM, S T Prasad <stprasad at stprasad.com> wrote:
>
>> Hi Alan,
>>
>> Been looking for something on the same lines and your question spurred me
>> into looking around.
>>
>> https://www.mysql.com/products/cluster/availability.html
>>
>> Seemed to be something I intend to try out.
>>
>> Of course the CRM would also have to be running via a Load Balancer on
>> seperate CRM Servers connecting to the mySQL Cluster.
>>
>>
>>
>> With best regards,
>>
>> S.T.Prasad
>>
>> http://www.vtigress.com
>> The Purr-fect mate for vTiger
>>
>> On Fri, May 12, 2017 at 6:13 PM, Alan Lord <alanslists at gmail.com> wrote:
>>
>>> We have a client who would really like a backup vtiger system which is
>>> 100% in-sync with the "master" so if the main server dies, they can switch
>>> to the backup without having suffered any data loss.
>>>
>>> I know 100% is almost certainly not possible, but am wondering if anyone
>>> has ever successfully setup a high-availablity backup scenario?
>>>
>>> TIA
>>>
>>> Al
>>>
>>> _______________________________________________
>>> http://www.vtiger.com/
>>>
>>
>>
>> _______________________________________________
>> http://www.vtiger.com/
>>
>
>
> _______________________________________________
> http://www.vtiger.com/
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.vtigercrm.com/pipermail/vtigercrm-developers/attachments/20170515/d0a7df84/attachment.html>


More information about the vtigercrm-developers mailing list