nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Architecture Application Server.

Julien.Fontaine

To continu with my questions about Architecture, i would want to know How you manage FailOver and HA with your application server (to run TaskSpace).

Do you use Apache and mod_jk or other technologies ?

Find more posts tagged with

Note:

If there are Accepted Answers, those will be shown by default. You can switch to 'All Replies' by selecting the tab below.

Button

Accepted answers

bacham2

Load balancing for the Content Server is usually done with the combination of DFC/Docbroker/Content Server. I don't know any customer using an external load balancer.

All comments

DCTM_Guru

We use hardware loadbalancer (eg F5). This it can handle failover at many levels (eg app server, content server, etc).

Julien.Fontaine

Hi Johnny,

You use F5 as loadbalance and failover for Application Server ?

You also use it as loadbalancer for Content Server ?

The F5 can be used for failover also ?

Is someone use a software solution ?

DCTM_Guru

I'm not a hardware guy, but yes, F5 can redirect to another server if a particular server in group is not up.

bacham2

Load balancing for the Content Server is usually done with the combination of DFC/Docbroker/Content Server. I don't know any customer using an external load balancer.

DCTM_Guru

Let me clarify, my previous statement. We use F5 more for failover for various components. I presume you could load balance the app server with F5 (since thats the first entry point and you need to configure sticky sessions between app server and content), but like I said, I'm not a hardware person, so I cant say for certain if "load balancing" is supported on F5.

Julien.Fontaine

Thanks Johnny,

I think F5 is a good point for hardware loadbalancer.

I also try to find some software failover / loadbalancer. Some idea ?

lastnitescurry

Well do not use Windows failover.... Had it on a project last year, because client IT team wanted it... but when I showed that the windows servers can be up but the content server dead..... folks not happy.

Use something that can monitior and manage indivdual compoents.... no sense in having hardware failover if the JMS is gone down.... it happens... btw... there are high availablity (monitioring?) scripts for that in the Content Server Admin guide

DCTM_Guru

Brian,

Curious, why was the CS dead? Was it b/c the CS service had to reboot - so therefore it really didnt "failover" in real-time?

lastnitescurry

Because I stopped it!!! I am a truely evil demon when it comes to testing!!!

But I did find a case last week where the repository was not running but docbroker and JMS were.... Oracle was on same box... TNS Listener had not started, hence no access to database, hence no reposiotry.....

In this case I had requested the server to reboot at day end... so it would be nice and fresh for me to work with the following morning.... was rather suprised when services not up.....

Julien.Fontaine

And do you already test a web server like nginx to answer failover and loadbalancing needs ?