RE: [load balancing] nodes causing lag

From: Massar, Marc <MMassarIZZATlinkpoint.com>
Date: Mon May 02 2005 - 13:23:27 EDT

Maybe a result of application pooling? We build a connection pool so the apps can get to the database and not have to worry about sessions to the db. Maybe as you add more nodes (and if you aren't pooling) you are building a bit of a backlog to your app layer or to your db and it appears as lag in the front.

        -----Original Message-----
        From: owner-lb-l@vegan.net on behalf of Chuck
        Sent: Mon 5/2/2005 9:04 AM
        To: lb-l@vegan.net
        Cc:
        Subject: [load balancing] nodes causing lag
        
        

        Doing some testing this weekend on a new 6400 (9.05) we saw some
        application delays (about 400-500 ms). Have a plain vanilla VIP with
        about 45 nodes (least connections member, no persistence, not
        mirroring, etc). Doing a pretty low BW/low user test I saw the delay
        (600 users 10Mbits). I dont think this is unique to the 6400 - we
        previously ran on 5100 - we have basically always seen this lag in our
        application - now we think we may know where it came from.
        
        Changed setup to use just one node and the delay went away (under same
        user/BW load). Through trial and error saw that as the # nodes
        increased, my delay increased. Jumped through numerous hoops - plugged
        nodes into BIGIP switch, used primary then secondary cat, added virutal
        interfaces, added ports, removed ports, had all nodes on one box,
        spread all nodes to 12 different boxes, went through fw, removed fw,
        turned off ssl, etc, etc ... In each permutation always the same - the
        more nodes - the more delay.
        
        Anyone have any thoughts on this? In my wildest imagination (and let
        me tell you it gets pretty wild) I cant imagine that simple LB to 45
        nodes introduces ANY latency - and especially not the 400 ms I am
        seeing. However - every test I have done points to that being the
        culprit.
        
        Open to suggestions for further tests, alternative conclusions, etc -
        would like to hear about anyone elses experience with pool-size, etc
        
        Not sure this matters - but our setup is RMI, with WebLogic
        ____________________
        The Load Balancing Mailing List
        Unsubscribe: mailto:majordomo@vegan.net?body=unsubscribe%20lb-l
        Archive: http://vegan.net/lb/archive
        LBDigest: http://lbdigest.com
        MRTG with SLB: http://vegan.net/MRTG
        Hosted by: http://www.tokkisystems.com
        
        
        

____________________
The Load Balancing Mailing List
Unsubscribe: mailto:majordomo@vegan.net?body=unsubscribe%20lb-l
Archive: http://vegan.net/lb/archive
LBDigest: http://lbdigest.com
MRTG with SLB: http://vegan.net/MRTG
Hosted by: http://www.tokkisystems.com

Received on Mon May 2 14:30:48 2005

This archive was generated by hypermail 2.1.8 : Mon May 02 2005 - 14:44:19 EDT