A redundant number of nodes make state dumps for both testnet and mainnet

[gmilescu]

A redundant number of nodes make state dumps for both testnet and mainnet.

[ND-571] created by None

[gmilescu]

https://docs.google.com/document/d/1_U3WnGqP_cvR58Uj252CxRO-7CIR9nG5m52TzF_Fpko/edit

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

https://github.com/PagodaPlatform/tf-near-node/pull/11

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

This PR is enabling state dumping for half of the nodes in every region

https://github.com/PagodaPlatform/tf-near-cluster/pull/3

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

node image changes https://github.com/PagodaPlatform/pkr-node/pull/4

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

this is ready, pending 1.36 RC

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

What I see is that 1 new testnet instances are dumping state parts:

• andreitestnet-canary-rpc-01-us-west1-a-2afd64c8
• nikrut-dump

And I see 0 mainnet instances dumping state parts, only nikurt-dump

andreitestnet-canary-rpc-01-us-west1-a-2afd64c8 seems to be doing perfectly fine, nice! ��

by 617a65ff58006300699436b5

[gmilescu]

the infra is prepared for both testnet and mainnet. I actually forgot to destroy the andreitestnet-canary-rpc-01-us-west1-a-2afd64c8 following successful tests.
Was planning wider fleet of nodes on 1.36 release. For now, I’ll add a new mainnet node dumping state.
Let me know if any objections to this

by 63b74c4d7cde7bff9d7ac618

[gmilescu]

andreimainnet-canary-rpc-01-us-west1-a-3af036aa

node added

by 63b74c4d7cde7bff9d7ac618