Maybe this is outside the scope, but what happen then I run a HTTP POST that are large, say 100MB. Will that be cached before sent to pool? Or does it have to do with http rechunk profile? What is the sizelimit on the http post?
I'm not sure there is a max size on the content of a POST unless you are limiting with ASM or an iRule. The data from the post is streamed as it arrives to the pool member unless again you want to inspect, which has to be done very carefully to avoid resource exhaustion.
When it comes to http idle server side profile. Does the big ip have the connection already to server ? What if there are no requests yet? Is it not waste of resources for big ip ? Lastly, how long does big ip hold the connection for the server in this profile ?
Jason, I get a lot of questions regarding why you use an IP address for the Forwarding L2 VS and not the MAC address. I haven't had time to set this up in my lab yet but my guess is that the F5 will ARP for the IP Address and obtain the MAC, add the MAC to its own ARP table and use that MAC address when forwarding its packets. Can you confirm that this is true? For all of you that are thinking about setting up a Forwarding L2 VS, since the IP address on the VS is the exact same as the end-server, remember to turn off the VS on the VLAN where the end-server resides using the VLAN And Tunnel Traffic setting on the VS and choose Disabled on.. and choose the VLAN where the end-server resides. If you do not do this the F5 will also ARP for the IP address that the end-server owns causing an IP conflict.
You are correct. The proper setup for the vlan group is to set the IP address of the self to the ip of node in the "back end" vlan, then disable that vlan on the virtual server to prevent the situation Philip describes. More details here: support.f5.com/kb/en-us/solutions/public/k/10/sol10371011.html
Queation 1- what is the major differnce between the l2 and layer3 forwarding virtual server Question 2-what is vlan group , why do we create it , why the extra configuration needed for L2 fwding VS? Question 3-performance vurtual server,performance http of performance l4 how do they increase performance?
1. l2 works on mac, l3 on ip 2. vlan group binds two vlans together so traffic is bridged between them 3. performance is about tradeoffs, you give up some functionality from standard vips by using the performance versions. For example, you can see the pros/cons of the fast http profile here: support.f5.com/csp/article/K8024
Quick and to the point. good job
is there ant role of self ip for the mentioned virtual servers types.
Maybe this is outside the scope, but what happen then I run a HTTP POST that are large, say 100MB. Will that be cached before sent to pool? Or does it have to do with http rechunk profile?
What is the sizelimit on the http post?
I'm not sure there is a max size on the content of a POST unless you are limiting with ASM or an iRule. The data from the post is streamed as it arrives to the pool member unless again you want to inspect, which has to be done very carefully to avoid resource exhaustion.
When it comes to http idle server side profile. Does the big ip have the connection already to server ? What if there are no requests yet? Is it not waste of resources for big ip ?
Lastly, how long does big ip hold the connection for the server in this profile ?
can you clarify the profile you are talking about?
For http and tcp both.
Jason, I get a lot of questions regarding why you use an IP address for the Forwarding L2 VS and not the MAC address. I haven't had time to set this up in my lab yet but my guess is that the F5 will ARP for the IP Address and obtain the MAC, add the MAC to its own ARP table and use that MAC address when forwarding its packets. Can you confirm that this is true?
For all of you that are thinking about setting up a Forwarding L2 VS, since the IP address on the VS is the exact same as the end-server, remember to turn off the VS on the VLAN where the end-server resides using the VLAN And Tunnel Traffic setting on the VS and choose Disabled on.. and choose the VLAN where the end-server resides. If you do not do this the F5 will also ARP for the IP address that the end-server owns causing an IP conflict.
You are correct. The proper setup for the vlan group is to set the IP address of the self to the ip of node in the "back end" vlan, then disable that vlan on the virtual server to prevent the situation Philip describes. More details here: support.f5.com/kb/en-us/solutions/public/k/10/sol10371011.html
Queation 1- what is the major differnce between the l2 and layer3 forwarding virtual server
Question 2-what is vlan group , why do we create it , why the extra configuration needed for L2 fwding VS?
Question 3-performance vurtual server,performance http of performance l4 how do they increase performance?
1. l2 works on mac, l3 on ip
2. vlan group binds two vlans together so traffic is bridged between them
3. performance is about tradeoffs, you give up some functionality from standard vips by using the performance versions. For example, you can see the pros/cons of the fast http profile here: support.f5.com/csp/article/K8024