-
Notifications
You must be signed in to change notification settings - Fork 1.6k
Fluent-bit on ECS Fargate - intermittent issue #6949
Replies: 1 comment · 10 replies
-
Can you give us the Fluent Bit log output?
May be sometimes the app is just causing the task to exit immediately and the app didn't produce any logs? Enable debug logging: https://github.com/aws/aws-for-fluent-bit/blob/mainline/troubleshooting/debugging.md#enable-debug-logging This will tell you if Fluent Bit even ingested any records. You should see a message like "update chunk with new size" |
Beta Was this translation helpful? Give feedback.
All reactions
-
what is the intermittent issue? |
Beta Was this translation helpful? Give feedback.
All reactions
-
@PettitWesley: I have elaborated the issue on above, Pasting here again. Checked with AWS team as well. but confirmed its issue with Fluent-bit level.For Up and running ECS task it is able to send logs to both the destination, but when there is new deployment and there is some issue for task to come up buz of some error at application code end and ECS is continue to spinning up new task to maintain the desired count. |
Beta Was this translation helpful? Give feedback.
All reactions
-
Please send me debug log output from Fluent Bit: https://github.com/aws/aws-for-fluent-bit/blob/mainline/troubleshooting/debugging.md#enable-debug-logging |
Beta Was this translation helpful? Give feedback.
All reactions
-
Hello @PettitWesley , Please find output.. this is the same where it didn't push any logs. 2023-03-14T19:18:45.903+05:30
|
Beta Was this translation helpful? Give feedback.
All reactions
-
@PettitWesley Can you check above please. |
Beta Was this translation helpful? Give feedback.
-
Hi Team,
My application is deployed on ECS Fargate based container. Where I have configured Fluent-bit for moving my ECS logs to OpenSearch and CloudWatch Log Group.
For Up and running ECS task it is able to send logs to both the destination, but when there is new deployment and there is some issue for task to come up buz of some error at application code end and ECS is continue to spinning up new task to maintain the desired count.
On this situation most of the time, Log router is not pushing any logs to destination but some time it is pushing. Kind of intermittent , Can someone please help with this
Below is my configuration:
--extra.com
[SERVICE]
flush 1
Grace 30
log_level debug
[OUTPUT]
Name cloudwatch
Match **
region us-east-1
log_key log
log_group_name ${LogGroup_Name}
log_stream_prefix ${env}-
auto_create_group true
[OUTPUT]
Name OpenSearch
Match **
Host ${ES_Host}
Port 443
Index graceit-${Service_Name}-dev-%Y-%m-%d
Type _doc
AWS_Auth On
AWS_Region us-east-1
tls On
Suppress_Type_Name On
retry_limit 3
Buffer_Size 600KB
Dockerfile
FROM amazon/aws-for-fluent-bit:latest
ADD extra.conf /fluent-bit/configs/extra.conf
ADD entrypoint.sh /entrypoint.sh
RUN chmod +x /entrypoint.sh
ENTRYPOINT ["/entrypoint.sh"]
@PettitWesley
Beta Was this translation helpful? Give feedback.
All reactions