[BUG] - chain_get_block RPC call results in request error: deadline has elapsed

[devendran-m]

Environment : Devnet
Bug/Issue reported by: Ihor from MAKE

Bug:

• The error occurs during rpc call with the method chain_get_block.
• Error details: Key: Node request failure, data: "request for signed block has failed: request error: deadline has elapsed."
• It consistently happens with chain_get_block, suggesting the error may not be related to request limits.
• This issue occurs even when the system isn’t under a high load (e.g., during isolated requests).
• Errors were observed after resuming data pulling from the updated Devnet.
• The error happens after a chain_get_era_info_by_switch_block request, with timeouts occurring 2 seconds after the request.

Additional info by the reporter:

I have more information:
When we received that error we made only one request during that second (however look at the 2 second duration!)
In total it looks like we aren’t making more than 80 reqs/sec

I looked at a little bit bigger sample. The same pattern repeats:

1. • Isolated chain_get_block request
2. • That happens after a chain_get_era_info_by_switch_block requests
3. • Timeouts after 2 seconds

It looks like the options are:

1. Issue in chain_get_block itself
2. Something related to chain_get_era_info_by_switch_block that causes the node (or the sidecar) to fail on the next request sometimes

This new sample has two errors that are 3 mins apart from each other, thus I don’t know if it can be related to an era end or something (I haven’t checked if there was an era end around that time on Devnet)