Forum Discussion
Kenny_Cheng_502
Nimbostratus
Mar 08, 2007big question on BIP-IP/IIS 6 web loadbalancing
Hi,
I'm evaluating the web loadbalancing product. One is F5 Bip-IP solution. I'm glad I can find this place.
I've been talking to the local sales/pre-sales technical guys. No one seems ...
Tech_Imp_40243
Mar 09, 2007Historic F5 Account
Hi Kenny,
Good questions.
On question 1:
In the section called Creating the HTTP health monitor in http://www.f5.com/solutions/deployment/iis_bigip9_dg.html:
There are 2 areas that you can send and receive strings to your application. This is pretty extensible because you can put in any valid HTTP command that you can post to the server. The Receive string will be what the BigIP looks for to determine if the application is up and running.
Additionally if you are looking to see if say 2 applications are working like the application and the database we can support a health monitor for that too. These are pretty extensive and I’d suggest that you take a look at this section in our application guide at:
https://tech.f5.com/home/bigip-next/manuals/bigip9_4/bigip9_4config/BIGIP_LTM_CONFIG_GD_9_4-13-1.htmlwp1201151
(The login is free)
On question 2:
This really is a 'depends' answer. You will need to look at striking a balance for how fast you need to determine if the server is up or vs. the amount of traffic / load you want to see on the server. On that same page as I mentioned above, you can set the interval for testing and then when you feel it has timed out.
We typically recommend the timeout be 3 x plus one second the amount of the interval test time; if you were doing a 20 second check the timeout interval would be 61 seconds. It is important to consider that on occasions a packet would be lost and you may not want to immediately mark the server down. So in the example above the monitor will try doing a check every 20 seconds; if in the span or 61 seconds it does not get a response it will mark the server down.
On question 3:
I'm not sure I understand this yet. Are you asking to have the user moved to another server if the one they were connected to is not working? If that is the case, this is typical BigIP behavior that when a server is down in the pool of servers we will mark that server as down and no longer route traffic to it. We will then balance any connections that had been going to now downed server to other servers in the load balanced pool.
BigIP transfers the session from one server to the other but the session can be a different thing. This really depends on the application and how it maintains session state. There are lots of options for this and at one end of the extreme the application can do state sharing across several servers to cover for this and at the other end of the extreme the customer may need to log back in to get a new session.
On question 4:
I'm not sure I understand this one yet. Can you fill me in with more details? Are you looking for running several different services on one server and one service could be down but the other service is up? If so, then BigIP is balancing the applications across the servers and we can easily support you there.
Best regards
Help guide the future of your DevCentral Community!
What tools do you use to collaborate? (1min - anonymous)Recent Discussions
Related Content
DevCentral Quicklinks
* Getting Started on DevCentral
* Community Guidelines
* Community Terms of Use / EULA
* Community Ranking Explained
* Community Resources
* Contact the DevCentral Team
* Update MFA on account.f5.com
Discover DevCentral Connects