You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Frequently, we'll see issues that are specific to differences in write times vs read times, and 'sar' data only provides a cumulative time for both read and write.
Linux and other OSes provide the read and write response times in addition to the cumulative times.
It would very beneficial if 'sar' data included this data as well.
I don't show any pending feature request of the same kind from what I could fine.
In the specific case of Linux, the data is provided in /proc/diskstats as:
Column Description
7 time spent reading (ms)
11 time spent writing (ms)
The text was updated successfully, but these errors were encountered:
I understand that "iostat" provides this. The issue is this distinction in missing in the historical data for 'sar', which makes diagnosing issues that occured in the past difficult that need this distinction.
In enterprise environments, the difference between read and write performance can vary significantly. The average of the two that 'sar' provides hides this from historical examination.
Frequently, we'll see issues that are specific to differences in write times vs read times, and 'sar' data only provides a cumulative time for both read and write.
Linux and other OSes provide the read and write response times in addition to the cumulative times.
It would very beneficial if 'sar' data included this data as well.
I don't show any pending feature request of the same kind from what I could fine.
In the specific case of Linux, the data is provided in /proc/diskstats as:
Column Description
7 time spent reading (ms)
11 time spent writing (ms)
The text was updated successfully, but these errors were encountered: