Add additional metrics to core engine

[masih]

Add metrics to report:

• Put latency normalised by number of multihashes (similar to the existing latency measure in Get ) -- so that we know how much time was spent exclusively on core layer to put indices.
• Average Put batch size of multihashes -- so that we can reason about optimisation of batching multihashes across multiple ads.
• Average size of keys and values -- so that we can reason about key-value store optimisation parameters.
• Average metadata size
• Removal rate by context ID and by provider -- so that we can estimate garbage size.

[masih]

@gammazero please edit the issue description as you see fit. Also please feel free to reduce the scope of metrics if you think some of them are YAGNI.

[gammazero]

Notes/questions on proposed metrics:

• Normalized Put latency is handled by recording the number of multihashes and elapsed time as an ingestion rate record. These records are collected by the telemetry service.
• Average Put batch size: I do not think we want to batch multihashes across multiple ads ever. Once we have the ad and all its entries, it is best to make sure that is committed before moving on to the next ad. If there is any interruption in indexer operation we want to only have to reindex the last complete ad. Especially when writing a CAR mirror.
• Is average size of keys and values also over last N samples? Do we also want to keep a ratio of keys to values?
• We currently track the overall percentage of removal ads. Tracking per provider would show which providers had the highest churn. It might makes sense for GC to track this because then we could track the number of multihashes removed, not just context IDs.