horizon ingest the historical data

Question

I'm running a new horizon v1.2 with ingest=true. It is connected to the stellar core with CATCHUP_COMPELTE enabled. (the status of the core is catching up 5%)

My question is what is the command to run the horizon to ingest all of the historical data? I used horizon serve the same as the previous versions of the horizon but doesn't ingest the old data.

I used horizon db reingest 1 29599089 and I get this error time="2020-05-11T14:19:55.542Z" level=error msg="Error in ingestion state machine" current_state="reingestHistoryRange(fromLedger=1, toLedger=29599089, force=false)" error="error processing ledger sequence=1: Error creating ledger reader: ledger not found" next_state=stop pid=1 service=expingest

Answer 1

My question is what is the command to run the horizon to ingest all of the historical data? I used horizon serve the same as the previous versions of the horizon but doesn't ingest the old data.

Currently, Horizon starts ingesting ledgers starting at the latest checkpoint and it won't start until this ledger is present in Stellar-Core.

time="2020-05-11T14:19:55.542Z" level=error msg="Error in ingestion state machine" current_state="reingestHistoryRange(fromLedger=1, toLedger=29599089, force=false)" error="error processing ledger sequence=1: Error creating ledger reader: ledger not found" next_state=stop pid=1 service=expingest

This error means that Horizon tried to ingest ledger 1 but it was not found in Stellar-Core database. You can try running:

select min(ledgerseq), max(ledgerseq) from ledgerheaders;

in Stellar-Core database to understand what's the first and the last ledger in it's DB (beware: gaps are possible).

It's possible that you have AUTOMATIC_MAINTENANCE_COUNT set in your Stellar-Core config so it returns older ledgers from time to time.