I am working on a HA solution for telemetry ingestion and curious if Grafana Alloy supports disk caching in case the export destination is offline etc.?
opentelemetry-collector/exporter/exporterhelper/README.md at main · open-telemetry/opentelemetry-collector (github.com)
Take a look at the documentation for prometheus.remote.write: prometheus.remote_write | Grafana Alloy documentation Does this cover what you’re looking for?
Thanks @claytoncornell, that helps with the metrics. I am looking for a global cache/queue mechanism if the export end points go down for logs and traces as well.
Hello! Prometheus.remote_write and loki.write have a WAL (note that the one in loki.write is still experimental).
Currently, there isn’t a WAL-like feature for otelcol components in Alloy. We are open to implementing Alloy equivalents to extensions such as the file storage one, but we haven’t prioritised it yet.