This content has been marked as final. Show 3 replies
In your example, maintenance and others have equal shares. That means that any available capacity that makes it to level 3 would be equally divided between the two, subject to there being demand from each.
So if mission-critical requests 20% of I/O and critical requests 30% of total (or 37.5% of remaining), 50% of original I/O capacity is still available at level 3.
Thanks for your response, as always it is helpful.
I did't get how did you get to the 37.5% figure, additionally I want to know is what percent of the remaining I/O requests represent the maximum amount allocate for the Maintenance group? will it be 25% as per your response that 50% is available to these groups and since both have equal share?
The 50% is the unconsumed I/O requests at level 1+level 2 , so what would be I/O priority for those unconsumed I/O when it get passed to the level 3 ( maintenance and other category)? I am specifically looking for I/O priority for the unconsumed I/O request which gets to level 3 category?
If you have 80% of capacity remaining at level 2, 37.5% of the remaining capacity equals 30% of the total capacity (37.5% * 80% = 30%)
In your example, the maintenance group will receive a maximum of 50% of what's remaining at level 3. So yes, if you have 50% remaining at level 3 and sufficient workload, maintenance will receive 25% of capacity.
I don't know how you're defining "priority" but in an IORM context priorities usually translate to levels in the plan: the lower-numbered levels will get priority access to I/O capacity before higher-numbered levels do.