Scroll airdrop triggers blob space saturation: What are the impacts?

Author: BREAD Source: X, @0xBreadguy Translation: Shan Ouba, Golden Finance

We have finally reached a state of blobs saturation again - which means they are no longer "almost free".

How did L2 (the second layer network) react to this? How much ETH was burned? Although the answers are not very exciting, let's discuss it anyway.

I also put together a dashboard to track the changes in L2's behavior, let's take a look.

What happened?

Since the Dencun upgrade (March 2024), blobs have been basically free most of the time.

For blobs to no longer be free, the number of blobs collectively released by L2 needs to reach 50% of the available blob space, which is the so-called "target". This is very similar to the mechanism of Gas after Ethereum's 1559 upgrade.

Yesterday, we hit this threshold:

Why did we hit this threshold? Scroll airdrop.

Scroll transaction volume increased dramatically when users claimed airdrops, meaning they needed to publish more batches of data to Ethereum blobs than usual.

This can be seen from their publishing behavior:

This sudden blob traffic caused L2s to collectively exceed the 50% threshold/upper limit, which means that the cost of interaction with Ethereum for all L2s using blobs has increased.

You can see the cost growth over the past 24 hours:

I have labeled "Price Discovery Latency" in the image above. This is the delay between blob space saturation and price discovery.

Due to some complex calculations, this process is longer than expected, and it is currently planned to be addressed in the next Pectra update (EIP-7762).

Combining all of the above:

How much ETH was burned?

In these hours, almost nothing was burned. A rough calculation is about $13,000 or so.

What do you want to observe?

I want to see if this airdrop surge is severe enough that L2 will change its behavior.

For example:

Will it switch from blobs to calldate for publishing data? Will it slow down or stop publishing altogether?

From what I have seen so far, neither has happened, but it is a good test practice. I also took the time to put together a Dune dashboard to more easily track visualizations showing these behavioral changes.