Increase deployment log limit to 10k lines for on-prem
complete
C
Cheerful Whitefish
Currently, our deployment logs are limited to 5000 lines, which causes them to be cut off. It would be beneficial to increase this limit to 10,000 lines, especially for on-premise versions, to ensure we can view the full logs without any truncation.
Log In
Prasad Satam
complete
Cheerful Whitefish - We are pleased to inform that we can increase the log limits on SMP to 10K or more, please see the instructions below and let us know if you have questions or follow ups.
Increasing the Log Line Limit in Harness Self-Managed Platform (SMP)
By default, newer installations of the Harness Self-Managed Platform (SMP) have
Protocol Buffers (protobuf)
enabled for the log service, which allows for a significantly higher log line limit compared to older setups.To confirm or enable this functionality, add the following to your Helm chart override YAML file:
yaml
platform:
log-service:
stream:
enableProtobuf: true
In most cases, this will be sufficient to support more than 10,000 log lines.
If your use case requires an even higher log limit, you can increase the Redis stream size used by the log service. Add or update the following configuration in your override YAML:
yaml
platform:
log-service:
stream:
enableProtobuf: true
config:
LOG_SERVICE_REDIS_MAX_STREAM_SIZE: "10000" # Increase this value as needed
This configuration allows for a larger Redis stream buffer, accommodating more log lines during deployments or CI/CD pipeline runs.
Note: These changes should be applied as part of your Helm upgrade or installation process. If you set a higher limit your redis instance may need more memory, otherwise it will crash. For most of the cases, 10k messages per stream is enough.
Shylaja Sundararajan
under review
Rohan Gupta
Merged in a post:
Add option to increase the amount of log lines per pipeline
D
Decent Zebra
We'd like the ability to increase the amount of logs that can be shown in the Pipeline's Console View. Ideally we'd like this to apply on a per pipeline basis since we know we only have 1 pipeline that needs extended logs while our other pipelines finish with less than 5k lines.