Eth indexer not require start block env variable #116
Conversation
ppca
commented
Jan 16, 2025
•
ppca
commented
- this will enable the eth indexer to start from block_number = latest_block from rpc call - 15 if the last_processed_block was empty or too far behind.
- Also added 2 metrics: eth sign request count and eth indexed block height
ppca
changed the title
| if latest_block - latest_handled_block < MAX_DELAYED_BLOCKS { | ||
| latest_handled_block + 1 | ||
| } else { | ||
| latest_block.saturating_sub(MAX_DELAYED_BLOCKS) + 1 |
There was a problem hiding this comment.
Great optimization!
Due to current rate limit failure rate, there is a small issue. I notice due to frequent rate limit hit, sometimes it can be 5-10 blocks delayed, so there is a small chance of >15 block delayed. If that happens, some block would be skipped.
Therefore, I suggest doing this check only once, when there is no last handled block.
There was a problem hiding this comment.
I'll increase the delay to 100 blocks. I also want this fix to help the case where a node goes back online after being offline for a while, in that case we don't want the node to start way back.
There was a problem hiding this comment.
@ppca I think we should apply the same logic (use the same function) that is used in the NEAR Indexer. I do not think there is a need to differentiate,
