Kyle Campbell

I would love to have web socket support @tj!

@rguske @zy286938348 I am also seeing this issue using containerd as my runtime. It came up ~16 days after my first deployment of `fluentbit 1.8.1` on a fresh `k8s 1.21`...

I *think* this is the RCA for me: https://github.com/fluent/fluent-bit/issues/3873 When I removed `Ignore_Older` the issues I was seeing went away quite quickly. Please tag me if you are able to...

Circling back to this and I'm no longer certain that `Ignore_Older` is the RCA for me & yes can confirm I'm using containerd. The issue still appears to occur from...

@srideok2 I don't think there is a permanent solution yet but I've talked to a few other people who had issues and we've had luck with reconfiguring our fluentbits to...

@Jonathan-w6d yes this helped for me. Also increasing the max buffer size. That said you'll still run into an issue with a memory leak and eventually need to restart the...

> I have all buffers set to 2M and it has never successfully flushed chunks from one specific node - the one that is logging stack traces (about 8k long)....

@vivek-rg nice! I tried implementing the strategy defined [here](https://aws.amazon.com/blogs/compute/optimizing-disk-usage-on-amazon-ecs/) but it didn't work out properly. I modified the stack but probably should have recreated. I also read something about this...

It would be also great if these Docker ECS builds didn't leave the lambda process hanging or post duplicate messages to slack. Using 0.9.13

This issue was solved in v5 but it's now back in v6.