Content
@
0 reply
0 recast
0 reaction
Parithosh
@parithosh
The Ethereum Dencun fork went live successfully earlier this week on Mainnet! We’ve been using our monitoring tool called Xatu to collect data on network health, find the first (of many) full blogposts here: https://ethpandaops.io/posts/dencun-fork-analysis/
1 reply
0 recast
4 reactions
Parithosh
@parithosh
CPU/Memory usage increase was marginal, the network usage also went up (but still under the 5Mbps range)
1 reply
0 recast
0 reaction
Parithosh
@parithosh
We used Blockprint and co-related it with Xatu data to see the spread of validators that went offline were spread evenly. This indicates that there was no client bug, but rather a few people forgetting to update their nodes.
1 reply
0 recast
0 reaction
Parithosh
@parithosh
Most blobs seem to arrive at the ~2.5s mark, with the p95 values on our sentries getting close to ~4s. The p95 value is concerning as 4s is the attestation deadline, but we might need to collect data over a longer duration.
1 reply
0 recast
0 reaction
Parithosh
@parithosh
It appears that the time difference from the first blob seen to the last blob seen in a slot is quite narrow. This is quite good news as a higher latency would indicate some routing issues on the p2p layer.
1 reply
0 recast
0 reaction
Parithosh
@parithosh
It appears that block propagation is ~10% worse when a slot has 6 blobs instead of 1 blob. This is still well under the 4s mark, but we are clearly getting closer to the attestation deadline and more analysis might be needed.
1 reply
0 recast
0 reaction
