[on-call] look into discrepancy between validator AS dashboards

[eelanagaraj]

Big discrepancy in validator AS completion rates for AnyblockAnalytics in the dashboards linked below:

• https://thecelo.com/groupDetail/0x69cd274C6AcBc08F664eD7B2D54aaB6615BC1d70
• https://celo.org/validators/explore
• https://rc1-blockscout.celo-testnet.org/address/0x66Be809e9ba5A6F57625b32506916db49a75D360/celo
• https://metabase.celo-networks-dev.org/public/dashboard/b0a27650-1d62-4645-81d7-26ff7546ff0d?date_filter=past4weeks~&validator_address=0x8C1Bf342693772A04bB223EA949f915ebf72521A
• https://cauldron.pretoriaresearchlab.io/attestations; Validator: AnyblockAnalytics; Attestation URL: https://attestation.celo.anyblock.tools

Outputs:

• understand what's causing the differences
• determine which dashboard(s) are closest to ground truth
• understand implication re: foundation votes, reach out to them if necessary
• communicate findings to Anyblock + other validators via discord/email

[eelanagaraj]

• Differences likely caused by a bug in the blockscout indexer's counter, will be investigated by Data Services in this ticket
• Raw blockscout logs confirm eksportisto data (i.e. AS metabase dashboards, thecelo)

With this in mind, seems even more important that the foundation voting's AS completion component is not based on https://celo.org/validators/explore until the blockscout bug is fixed cc @codyborn @arthurgousset

[0xarthurxyz]

Thanks for the detailed overview 💯 🙌
I don't fully understand how the "foundation voting" program works.

I.e. is it programmatic or does someone from the Foundation look at the dashboard (https://celo.org/validators/explore) to make manual adjustments to the voting program.

If someone is manually doing this, we could simply suggest they use the Metabase dashboard instead like you said.