Comments on: Load Balancing Performance Metrics 101 http://lbdigest.com/2009/05/20/load-balancing-performance-metrics-101/ Server Load Balancing Articles and News Wed, 11 Aug 2010 14:07:20 +0000 hourly 1 http://wordpress.org/?v=3.0.1 By: HO http://lbdigest.com/2009/05/20/load-balancing-performance-metrics-101/comment-page-1/#comment-1097 HO Mon, 05 Oct 2009 21:23:32 +0000 http://lbdigest.com/?p=372#comment-1097 Did your content get truncated? The "Connection Rate versus Request Rate" section appears to have been cut off. In your section on "Layer-4 versus Layer-7", one additional note is that the traffic mix, that is ratio of Layer-4 (simple routing) to Layer-7 (packet inspection) traffic, and the longevity of these transactions can also impact the strict TPS rate that you can get out of a particular appliance. On a side note, I have been in situations where the production network was multi-server behind a load-balancer with SSL termination and re-encryption to the back-end servers. All well and good, but the development network was a single server sans load-balancer, sans SSL... And they were trying to compare TPS of the two networks... I was unable to convice them that offloading to a CDN won't make as much of a difference as they hoped. Did your content get truncated? The “Connection Rate versus Request Rate” section appears to have been cut off.

In your section on “Layer-4 versus Layer-7″, one additional note is that the traffic mix, that is ratio of Layer-4 (simple routing) to Layer-7 (packet inspection) traffic, and the longevity of these transactions can also impact the strict TPS rate that you can get out of a particular appliance.

On a side note, I have been in situations where the production network was multi-server behind a load-balancer with SSL termination and re-encryption to the back-end servers. All well and good, but the development network was a single server sans load-balancer, sans SSL… And they were trying to compare TPS of the two networks… I was unable to convice them that offloading to a CDN won’t make as much of a difference as they hoped.

]]> By: stine http://lbdigest.com/2009/05/20/load-balancing-performance-metrics-101/comment-page-1/#comment-1082 stine Mon, 01 Jun 2009 19:39:20 +0000 http://lbdigest.com/?p=372#comment-1082 Did your content get truncated? The "Connection Rate versus Request Rate" section appears to have been cut off. In your section on "Layer-4 versus Layer-7", one additional note is that the traffic mix, that is ratio of Layer-4 (simple routing) to Layer-7 (packet inspection) traffic, and the longevity of these transactions can also impact the strict TPS rate that you can get out of a particular appliance. On a side note, I have been in situations where the production network was multi-server behind a load-balancer with SSL termination and re-encryption to the back-end servers. All well and good, but the development network was a single server sans load-balancer, sans SSL... And they were trying to compare TPS of the two networks... I was unable to convice them that offloading to a CDN won't make as much of a difference as they hoped. Did your content get truncated? The “Connection Rate versus Request Rate” section appears to have been cut off.

In your section on “Layer-4 versus Layer-7″, one additional note is that the traffic mix, that is ratio of Layer-4 (simple routing) to Layer-7 (packet inspection) traffic, and the longevity of these transactions can also impact the strict TPS rate that you can get out of a particular appliance.

On a side note, I have been in situations where the production network was multi-server behind a load-balancer with SSL termination and re-encryption to the back-end servers. All well and good, but the development network was a single server sans load-balancer, sans SSL… And they were trying to compare TPS of the two networks… I was unable to convice them that offloading to a CDN won’t make as much of a difference as they hoped.

]]>