44433f38be
also remove the continuation support from the queue, otherwise we may end up sequential scanning the entire table to get a few rows at the end. then, in the core, instead of looping both to get a big enough batch inside of the queue, as well as outside of it to ensure we consume the whole queue, just get a single batch at a time. also, make the queue size configurable because we'll need to do some tuning in production. Change-Id: If1a997c6012898056ace89366a847c4cb141a025 |
||
|---|---|---|
| .. | ||
| live | ||
| reportedrollup | ||
| rollup | ||
| tally | ||
| billing_test.go | ||
| bucketstats.go | ||
| bucketusage.go | ||
| common.go | ||
| db_test.go | ||
| db.go | ||
| projectusage_test.go | ||
| projectusage.go | ||