Dynamic GTM load balancing for generic hosts?
Can a GTM be to used to dynamically load balance between generic host servers. [I don't have an LTM] I see that the big3d agent is needed to collect statics for dynamic load balancing, and I can see how this works when load balancing across different F5 appliances. But I want to load balance between generic host servers. I see I can do static load balancing, e.g. round robin, but just wondered if it was possible to do anything more sophisticated? Is dynamic load balancing applicable in this instance, the documentation is unclear. Thanks
LTM with DNS - logging query answers DNS_RESPONSE clientside
Hi All We have our DNS services behind LTM VIPs. We have the DNS license and are using DNS_QUERY and DNS_RESPONSE events for logging queries and answers. We are not using Express, BIGIP Bind, nor GTM configurations - straight LB work. Last week I was investigating some optimizations and wanted to add Answer header information, specifically the truncate flag. This will allow us to gather some stats on the amount of UDP to TCP based queries occuring. I added[DNS::headertc] to the log message on DNS_RESPONSE and proceeded to test from a client system using DIG for a test SRV record that exceeded 512bytes using UDP. DIG did the expected things and sent a UDP query, got a response with truncate on, and re-queried with TCP. All good. However, when I looked at the logs,[DNS::headertc] was always returning 0. From network trace, I see client to VIP, SNAT to service, service to SNAT, VIP to client. The client side traffic is as expected - no EDNS0, buffer set to 512bytes, etc. As well the answer in UDP shows the truncate flag set to 1. Further examination showed that while the client had specified no EDNS0, buffer 512bytes, etc., the SNAT to DNS service traffic showed EDNS0 on and buffer of 4096bytes. I asked our F5 account rep if he had any insights and he agreed that DNS_RESPONSE seems to be pulling from server-side. I tried using [clientside {DNS::header tc}] in the logging statement, but got the same results - truncate still shows 0. Questions: 1. Is there a way to tell LTM to respect the client settings for the server side communications? 2. Can I get the client-side info in the DNS_RESPONSE event? Thanks
Uneven balancing - ratio (member)
Hi everybody, So here is my situation, as you can see here: I have a pool (Pool_Extranet) which contains 2 nodes ExtranetB and ExtranetC where B is a physical server and C a virtual server. The thing is B is much more powerful than C and our customers would like to do a 75/25 balancing but as you can see it's not the case at all, 95% of the connection are on C. We tried to put the pool on ratio (member) and configure the ratio but it changes anything, same goes for dynamic ratio (member) but change anything too... We thought about rebooting ExtranetC to "move" the connections on ExtranetB but it's our last solution and wanted to try the rest before doing it. Any idea? Thanks in advance.
Load balancing using an API
Hello team, We have a bunch of hosts running behind F5. Every host is running few services. One particular service is capable of providing free memory information through the API we developed: GET http://hostname/myservice/usageAPI Response: { "freeMemory": 369959592 } Is it possible to consume this API in F5, and load balance accordingly? E.g. If freeMemory is less than threshold, than no request should be sent to that host for the time being. After sometime, when freeMemory is above the threshold value, then F5 should redirect request to that host. How to load balance in F5 through such API? Note that we don't want to mark server/host status Up and Down. We just want to make sure that particular service has enough memory to take up the next memory intensive request. We know Dynamic Ratio Load Balancing but that considers the overall health of the host. We want to load balance based on status of one service out of few other services running on the host.
What is the equivalent of Cisco ACE Least-Bandwidth Predictor Method on LTM?
Hi, I have a customer who is migrating from ACE to LTM and one of their services uses the Least-Bandwidth Predictor Method: http://www.cisco.com/c/en/us/td/docs/app_ntwk_services/data_center_app_services/ace_appliances/vA3_1_0/configuration/slb/guide/slbgd/rsfarms.htmlwp1072278 AFAIK there is no equivalent on LTM as it doesn't collect metrics on bandwith, only on connections. I am starting to resign myself on having to use Predictive LB Method. Any other ideas? ThanksLTM - Least Connections loadbalancing method
Hello, Lets assume that we have 2 pools with different name having both the same 2 members. On both pools least connections load balancing method have been chosen based on member. By using irules we specify which pool is gonna be chosen based on the [HTTP::host] content. My question is how the F5 LTM will handle these requests ? Would it be the same to have just one pool with those 2 members ??? In other words by using 2 pools the F5 runs 2 instances for the least connection LB method (member) or it counts the connections in total ? Based on the same scenario in case we use least connection LB method (NODE) do we expect the same behaviour ?? Thanks
Radius Authentication and Accounting Persistence Across Servers
Hey All, We're looking to load balance radius authentication requests. I've been looking around DevCentral and found the following threads which have been helpful: Load Balance Cisco ISE Servers Radius Load Balancing Persistence We'll be utilizing the Radius profile to ensure that authentication requests persist to the same server based off of the Call-Station-ID. I have a few additional questions though. Is it possible to ensure that the accounting information goes to the same server in which the authentication transactions happened? BrianQuestion on Priority Group Activation
Hi, I want to make my virtual server with 9 pool member automatically disabled when four of its pool member are down. Can I achieve this with below settings : 1. Put all the pool members to the same priority group for example 5 2. Under Priority Group Activation, I would select 6 viz., traffic should be processed by the pool members of group 5 till the pool have 6 minimum active members failing which the group shall not process the traffic. Now, as all the pool members belong to same priority group 5 and when PGA conditions fails would the virtual server would be down as there are no more pool members to accept the traffic ?? Please provide your inputs. Thanks, MSK
