Hi team, we are sending logs from a remote VM to self-hosted signoz (Signoz otel collector )in our cluster via otel collector running in VM receiver filelog and otlp exporter. We want to have this setup in production. but we have reliability concerns like if the VM otel collector is down or signoz otel collector is down. we will miss logs. Signoz team please help with a detailed explanation for this usecase this would be really helpful and it is only blocker for us. @Nitish @prasanth

Interesting! Curious to know why did you self-host signoz rather than using SigNoz cloud?

You will miss some logs if your VM otel collector is down and logs are being read from end of file, but if the signoz-otel-collector is down and comes back in some amount of time it should be fine as there are retry mechanisms in VM otel collector.

@Nitish We have a compliance issue with sending any data outside the cluster, we were using Elastic Cloud earlier. now have to move to a self-hosted solution.

maintain any pointer to read data from where it is left

No it doesn’t maintain any pointer. you can configure retry mechanism and queue here, it depends on how much resources you can allocate to make it retry-able here are the docs https://github.com/open-telemetry/opentelemetry-collector/blob/main/exporter/exporterhelper/README.md

thanks @nitya-signoz, checking above

Good to hear