-
Bug
-
Resolution: Cannot Reproduce
-
Critical
-
None
-
Logging 5.7.5
-
False
-
None
-
False
-
NEW
-
NEW
-
-
-
Proposed
Description of problem:
Collector pod deployed by the cluster-logging operator is using 1.3 cores of cpu on SNO DUs
Version-Release number of selected component (if applicable):
cluster-logging.v5.7.5 Tested on: 4.13.4 4.14.0-ec.4 4.14.0-0.ci-2023-09-05-120503
How reproducible:
1. Configure SNO with DU profile. 2. Deploy cluster-logging operator and configure clusterLogging.yaml and clusterLogForwarding.yaml 3. Deploy Telco workloads and gather resource usage metrics
Actual results:
CPU Usage in Cores:
| Iteration | 4.13.4 | 4.14.0_ec.4 | 4.14.0-0.ci-2023-09-05-120503 |
| 1 | 0.7119231194 | 0.7413879821 | 1.314202569 |
| 2 | 0.6751107603 | 0.7218595507 | 1.319234136 |
| 3 | 0.6984978924 | 0.7775478173 | 1.306324442 |
| 4 | 0.7541069158 | 0.7327866483 | 1.329228295 |
| 5 | 0.7484353611 | 0.7873075163 | 1.300583959 |
| 6 | 0.7753512426 | 0.8063833829 | 1.299729106 |
| Avg | 0.7272375486 | 0.76 | 1.31 |
CPU usage is high on all three releases. There is also an increase of ~0.58 cores between 4.14.0-0.ci-2023-09-05-120503 and 4.13.4.
Expected results:
Additional info:
clusterLogging.yaml
---
apiVersion: logging.openshift.io/v1
kind: ClusterLogging
metadata:
name: instance
namespace: openshift-logging
spec:
managementState: "Managed"
curation:
type: "curator"
curator:
schedule: "30 3 * * *"
collection:
logs:
type: "vector"
vector: {}
clusterLogForwarder.yaml
---
apiVersion: "logging.openshift.io/v1"
kind: ClusterLogForwarder
metadata:
name: instance
namespace: openshift-logging
spec:
outputs:
- type: "kafka"
name: kafka-open
url: tcp://e28-h29-r750.alias.bos.scalelab.redhat.com:9092/e28-h37-r750
pipelines:
- name: audit-logs
inputRefs:
- audit
outputRefs:
- kafka-open
- name: infrastructure-logs
inputRefs:
- infrastructure
outputRefs:
- kafka-open
