You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@yunikorn.apache.org by cc...@apache.org on 2022/03/23 15:21:50 UTC
[incubator-yunikorn-site] branch master updated: [YUNIKORN-1130] Update website documentation to reflect resource units (#141)
This is an automated email from the ASF dual-hosted git repository.
ccondit pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/incubator-yunikorn-site.git
The following commit(s) were added to refs/heads/master by this push:
new 5238998 [YUNIKORN-1130] Update website documentation to reflect resource units (#141)
5238998 is described below
commit 52389988ffbb01881598067195489560000192bc
Author: Craig Condit <cc...@apache.org>
AuthorDate: Wed Mar 23 10:21:38 2022 -0500
[YUNIKORN-1130] Update website documentation to reflect resource units (#141)
Closes: #141
---
docs/user_guide/queue_config.md | 35 +++++++++++++++++++---------------
docs/user_guide/resource_quota_mgmt.md | 31 +++++++++++++++---------------
2 files changed, 35 insertions(+), 31 deletions(-)
diff --git a/docs/user_guide/queue_config.md b/docs/user_guide/queue_config.md
index ef66b8a..7dd86ac 100644
--- a/docs/user_guide/queue_config.md
+++ b/docs/user_guide/queue_config.md
@@ -182,9 +182,9 @@ partitions:
parent: true
resources:
guaranteed:
- {memory: 1000, vcore: 10}
+ {memory: 1G, vcore: 10}
max:
- {memory: 10000, vcore: 100}
+ {memory: 10G, vcore: 100}
```
### Placement rules
@@ -235,8 +235,8 @@ limits:
- <group name>
maxapplications: <1..maxint>
maxresources:
- <resource name 1>: <0..maxint>
- <resource name 2>: <0..maxint>
+ <resource name 1>: <0..maxint>[suffix]
+ <resource name 2>: <0..maxint>[suffix]
```
Limits are applied recursively in the case of a queue limit.
@@ -311,13 +311,18 @@ Basic `resources` yaml entry:
```yaml
resources:
guaranteed:
- <resource name 1>: <0..maxint>
- <resource name 2>: <0..maxint>
+ <resource name 1>: <0..maxint>[suffix]
+ <resource name 2>: <0..maxint>[suffix]
max:
- <resource name 1>: <0..maxint>
- <resource name 2>: <0..maxint>
+ <resource name 1>: <0..maxint>[suffix]
+ <resource name 2>: <0..maxint>[suffix]
```
-Resources that are not specified in the list are not limited, for max resources, or guaranteed in the case of guaranteed resources.
+Resources that are not specified in the list are not limited, for max resources, or guaranteed in the case of guaranteed resources.
+
+An optional suffix may be specified for resource quantities. Valid suffixes are `k`, `M`, `G`, `T`, `P`, and `E` for SI powers of 10,
+and `Ki`, `Mi`, `Gi`, `Ti`, `Pi`, and `Ei` for SI powers of 2. Additionally, resources of type `vcore` may have a suffix of `m` to indicate millicores. For example, `500m` is 50% of a vcore. Units of type `memory` are interpreted in bytes by default. All other resource types have no designated base unit.
+
+Note that this is a behavioral change as of YuniKorn 1.0. Prior versions interpreted `memory` as units of 1000000 (1 million) bytes and `vcore` as millicores.
### Child Template
@@ -346,19 +351,19 @@ As an example:
application.sort.policy: stateaware
resources:
guaranteed:
- vcore: 1000
- memory: 1000
+ vcore: 1
+ memory: 1G
max:
- vcore: 20000
- memory: 600000
+ vcore: 20
+ memory: 600G
queues:
- name: parent
parent: true
childtemplate:
resources:
max:
- vcore: 21000
- memory: 610000
+ vcore: 21
+ memory: 610G
- name: notemplate
parent: true
```
diff --git a/docs/user_guide/resource_quota_mgmt.md b/docs/user_guide/resource_quota_mgmt.md
index 8890fdb..cc07f48 100644
--- a/docs/user_guide/resource_quota_mgmt.md
+++ b/docs/user_guide/resource_quota_mgmt.md
@@ -87,13 +87,12 @@ spec:
hugepages-1Gi: "1"
```
The above specification will set pod resources request for scheduling in YuniKorn to:
-* _vcore_ -> 250
-* _memory_ -> 1074
+* _vcore_ -> 250m
+* _memory_ -> 1073741824
* _hugepages-1Gi_ -> 1
Two remarks:
-Multiple container specifications will be aggregated into one total pod resource request automatically.
-All memory is reported in MB with unit conversions applied where needed.
+Multiple container specifications will be aggregated into one total pod resource request automatically. All memory is reported in bytes.
In the case that static queue definitions are used for a queue there is no limit on the type of resource that can be specified in a quota.
Quota annotations on namespaces, used as part of the automatic queue creation, are limited to the equivalent _cpu_ and _memory_ resources.
@@ -150,27 +149,27 @@ partitions:
- name: advertisement
resources:
guaranteed:
- memory: 500000
- vcore: 50000
+ memory: 500G
+ vcore: 50
max:
- memory: 800000
- vcore: 80000
+ memory: 800G
+ vcore: 80
- name: search
resources:
guaranteed:
- memory: 400000
- vcore: 40000
+ memory: 400G
+ vcore: 40
max:
- memory: 600000
- vcore: 60000
+ memory: 600G
+ vcore: 60
- name: sandbox
resources:
guaranteed:
- memory: 100000
- vcore: 10000
+ memory: 100G
+ vcore: 10
max:
- memory: 100000
- vcore: 10000
+ memory: 100G
+ vcore: 10
```
### Run a workload