The commands per port is by physical port. Paths in those two states must be recovered manually with chpath for Disabled paths or with cfgmgr for Missing paths. Find all posts by h foorsa. When using a switch change the fibre channel adapter to enable fast fail and dynamic tracking: Another feature of the recent improvements to the AIX MPIO software is the expansion of the output of the lsmpio command to include error counts per adapter and per remote port, allowing the SAN administrator to better deduce the source of intermittent, recurring errors. Indicates the maximum time spent by a transfer request in the wait queue.
|Date Added:||14 October 2018|
|File Size:||48.5 Mb|
|Operating Systems:||Windows NT/2000/XP/2003/2003/7/8/10 MacOS 10/X|
|Price:||Free* [*Free Regsitration Required]|
IBM AIX MPIO: Best practices and considerations
AIX Recommended Settings
SAN administrators may use several configuration techniques to maximize redundancy. Indicates the average time spent by a transfer request in the wait queue.
If kpio two switches, there is no single point of failure. Please repeat this for all fsc devices. For example, if the fiber connected to a particular AIX FC port is pulled out, all currently used paths originating at that port move to the Failed state. Generally, the hdisk attributes should be left at their default values, especially the attributes that are not mentioned in this article.
There are several possible combinations of using VIOS with Pure Storage, and the configurations differ depending on which one you are using. So nonzero counts, especially large nonzero counts, are more interesting than a small or zero error count. Note that if the error count for the remote ports is zero or very small, it may indicate that the port is working fine or it may indicate that the port is not currently being used. You can now continue adding your subsequent zoned paths and LUN’s.
The lsmpio command in older technology levels of AIX supports these two flags. Best practice is to make the queue depth the same in both the server and client partition. Running the lsmpio -az command resets all of the above counters to zero. Subscribe and stay informed about: For example, a marginal, constantly bouncing link in the SAN fabric might lead to this type of error recovery, resulting in extreme performance degradation.
IBM AIX multipath I/O (MPIO) resiliency and problem determination
However this error detection and recovery might take some mio, and that delay might impact applications. Indicates the maximum time spent by a transfer request in the wait queue. This setting causes the path to be disabled.
These metrics are not applicable for tapes. Typically, users configure the SAN fabric with more than one path between the computer and the storage device.
However, the cost of setting a lower health check interval far outweighs the benefits. Default is in milliseconds. If multiple disks were to encounter similar issues at the same time, the consequences for applications might be severe. The optimal configuration for a device having four paths on AIX is to use four iax paths to the storage subsystem with a 1: View Public Profile for h foorsa.
Consider using virtual scsi devices for boot to eliminate this. Likewise, for such failures it mlio also easy for the AIX MPIO software to detect and react to the failure by discontinuing the use of any path that uses the failed component. If you are running this version, run the following to clear the reservation from AIX before proceeding: