Export extra metrics for conflation, blob and aggregation

[jonesho]

Description

These extra metrics for conflation, blob and aggregation were requested by the prover team in #3128 :

For block

• Avg tx data on L1 (size of the average uncompressed data of each tx: the size of the entire uncompressed L2 block data divided by the number of tx in the block)
• Avg compressed tx data on L1 (size of the average compressed data of each tx: the size of the entire compressed L2 block data divided by the number of tx in the block)
• Avg compression ratio of tx data on L1 (1.0 - "avg post-compressed tx data size" / "avg pre-compressed tx data size") Duplicated by the blob compression ratio metric (linea_blob_block_compression_ratio)

For conflation

• L1 data (size of the uncompressed L2 data that would be posted on L1 via blob transaction)
• L1 data compressed (size of the compressed L2 data that would be posted on L1 via blob transaction)
• The limits (which already include L1 data) Can be skipped given that traces limits would be removed soon
• Total gas used

For blobs

• Number of conflation Done
• Number of blocks Done
• Size of the uncompressed data
• Size of the compressed data
• Total gas used

For aggregation

• Number of bridge messages (L1 and L2) Not a priority, can be revisited later and it can be on prover side to avoid the use of coordinator database
• Total gas used Not a priority, can be revisited later
• Number of conflation Done
• Number of block Done
• Number of blobs Done

Motivation

Useful for prover and arithmetization team to perform data analysis

Tasks

• Export histogram metric about total gas used in each conflation
• Export histogram metric about size of the uncompressed data in each conflation
• Export histogram metric about size of the compressed data in each conflation
• Export histogram metric about average uncompressed transaction data size in each conflation
• Export histogram metric about average compressed transaction data size in each conflation
• Export histogram metric about total gas used in each blob
• Export histogram metric about size of the uncompressed data in each blob
• Export histogram metric about size of the compressed data in each blob
• Export metrics about total gas used in each aggregation
• Export metrics about number of bridge messages in each aggregation

Acceptance criteria

• All metrics above are exported

Risks

• N/A

Remember to

• Add the documentation label in case there is an impact on the documentation
• Add priority and team labels
• Add Task for updating the Runbook or adding/updating existing metrics and alerts.

[jonesho]

@dfox-consensys FYI, left some questions next to the metrics, and below was the Alex's responses for my questions previously:

I just wonder "Total gas" in each "conflation, blobs, aggregation" means "the sum of L2 used gas from all included blocks" right?

Yes

what does "L1 data", "L1 data compressed", "The limits" mean for each conflation?

L1 data: The byte size of the L1 data summed for each block etc..
L1 data compressed: the same thing but for compressed data
Limits: The module limits

[dfox-consensys]

The metrics I am most interested in is the avg data size of a L2 tx that gets posted to L1, pre-compression, post-compression, and then the avg compression ratio