Right now the cursor has to be scraped from the log lines from plc-replica.
I plan on writing my own "microservice" (in my monolithic server binary :) for synchronous handle and service lookup.
The initial backfill is done by plc-replica, then the database is exported as JSON lines (going to file an issue for this), and then it continues by using the plc-replica cursor with /export/stream. I'll then continue running plc-replica as a source to recover from if https://plc.directory/export/stream returns OutdatedCursor to my application.
Right now the cursor has to be scraped from the log lines from
plc-replica.I plan on writing my own "microservice" (in my monolithic server binary :) for synchronous handle and service lookup.
The initial backfill is done by
plc-replica, then the database is exported as JSON lines (going to file an issue for this), and then it continues by using theplc-replicacursor with/export/stream. I'll then continue runningplc-replicaas a source to recover from ifhttps://plc.directory/export/streamreturnsOutdatedCursorto my application.