Monitoring SAN subnet delays in VMware vSphere 4.1
In this and the next posts we will try to reveal some interesting technological moments from those that we regularly meet in our work. This is partly prompted by the fact that the features of tariffing, the allocation and distribution of capacity are prosaic, and Habr has already seen a lot of posts on this topic. And real experience in solving some small, but important moments can be of great use. So spread the word!
')
The version of the VMware vSphere 4.1 product has a very useful feature for tracking delays and “sagging” of SAN-subnets based on Fiber Channel and Ethernet (NFS and iSCSI). For various reasons, from time to time, subnets of the general storage network cause delays in the passage of traffic between hypervisor hosts (Host Bus adapters) and the storage storage system processor.
In previous versions of vSphere, in order to deal with delays, the administrator could view the iSCSI command queue on the storage controller. But it was impossible to clearly see at which stage (the iSCSI-team) the passage of traffic through the switch there was a delay, which imposed certain restrictions on the work.
In vSphere 4.1, a latency tracking tool on Host Bus hypervisor adapters has been introduced. The administrator now has a tool for analyzing latency on local storage system Host Bus adapters, soft-HBA (iSCSI software adapter) software adapters and FiberChannel hardware adapters.
The tool allows you to track delays on each block disk device having a logical LUN address, logical unit number. Thus, the administrator receives a convenient means to control delays and promptly eliminate problems in data transfer between devices, which favorably affects the performance of the entire storage network.
')
The version of the VMware vSphere 4.1 product has a very useful feature for tracking delays and “sagging” of SAN-subnets based on Fiber Channel and Ethernet (NFS and iSCSI). For various reasons, from time to time, subnets of the general storage network cause delays in the passage of traffic between hypervisor hosts (Host Bus adapters) and the storage storage system processor.
In previous versions of vSphere, in order to deal with delays, the administrator could view the iSCSI command queue on the storage controller. But it was impossible to clearly see at which stage (the iSCSI-team) the passage of traffic through the switch there was a delay, which imposed certain restrictions on the work.
In vSphere 4.1, a latency tracking tool on Host Bus hypervisor adapters has been introduced. The administrator now has a tool for analyzing latency on local storage system Host Bus adapters, soft-HBA (iSCSI software adapter) software adapters and FiberChannel hardware adapters.
The tool allows you to track delays on each block disk device having a logical LUN address, logical unit number. Thus, the administrator receives a convenient means to control delays and promptly eliminate problems in data transfer between devices, which favorably affects the performance of the entire storage network.
Source: https://habr.com/ru/post/113457/
All Articles