Hi Polygon team,
I hope it’s the right place to ask a question about setting up an archive Polygon node (if not feel free to re-route me to wherever I can ask).
Running this with Kubernetes, PV set to a max of 16TB storage, deployment is having a max limit of 20GB ram and 4 core cpu limit.
Running a pod from a Dockerfile starting from Ubuntu 20.04.4 (Focal Fossa)
Building Heimdall from https://github.com/maticnetwork/heimdall/archive/refs/tags/v0.2.9.zip (unzip + make build)
Run Heimdall related with
heimdall start –home my_home > my_heimdall_log_path.log & heimdall rest-server –home my_home > my_rest_log_path.log &
sleep 50 (so I’m sure it’s all started)
bridge start –all –home my_home > my_bridge_log_path.log
I’m also starting off a snapshot https://matic-blockchain-snapshots.s3-accelerate.amazonaws.com/matic-mainnet/heimdall-snapshot-2022-08-02.tar.gz
It syncs OK and a curl to
catching_up: false at which point I start an Erigon node as well.
At some point I can see my bridge logs show this error:
E[2022-08-05|06:58:43.806] Error fetching block header from mainchain while calculating next span seed module=bor error="Post \http://localhost:9545\: dial tcp [::1]:9545: connect: connection refused" E[2022-08-05|06:58:43.807] Error fetching proposers module=span service=processor error="Error while fetching data from url: http://0.0.0.0:1317/bor/prepare-next-span?chain_id=137&proposer=0xc19d41fa1f612983322ee8b5b68a71f3bdbf3320&span_id=4930&start_block=31545856, status: 400"
I made sure 9545 is exposed in Dockerfile as well as from Kubernetes deployment and service as well.
A bit of research done and I stumbled upon something that seems similar:
My questions are:
How can I fix this error? A manual stop/restart of the bridge does not sound like an option for long-term use.
Does this affect my syncing (both for Heimdall and Erigon) OR the syncing speed in any way?
I got another question as well – as I was running the same setup but with Bor in the past and now switched to Erigon I’m still waiting for Heimdall to be in
catching_up: false state and then start Erigon as it was with Bor.
Is this still the case or I can start Erigon even if Heimdall is still catching up?
Thanks in advance!