File tree Expand file tree Collapse file tree 1 file changed +7
-5
lines changed
content/en/docs/concepts/cluster-administration Expand file tree Collapse file tree 1 file changed +7
-5
lines changed Original file line number Diff line number Diff line change @@ -96,13 +96,15 @@ Pods. This means that an ill-behaved Pod that floods the API server with
96
96
requests cannot prevent leader election or actions by the built-in controllers
97
97
from succeeding.
98
98
99
- ### Request Width
99
+ ### Seats Occupied by a Request
100
100
101
101
The above description of concurrency management is the baseline story.
102
- In it, all requests have equal "width": each takes up one _ unit of
103
- concurrency_ , which we call a "seat" because this is similar to how
104
- each passenger takes up one of the fixed supply of seats on a train or
105
- aircraft.
102
+ In it, requests have different durations but are counted equally at
103
+ any given moment when comparing against a priority level's concurrency
104
+ limit. In the baseline story, each request occupies one unit of
105
+ concurrency. The word "seat" is used to mean one unit of concurrency,
106
+ inspired by the way each passenger on a train or aircraft takes up one
107
+ of the fixed supply of seats.
106
108
107
109
But some requests take up more than one seat. Some of these are ** list**
108
110
requests that the server estimates will return a large number of
You can’t perform that action at this time.
0 commit comments