I was wondering, has anyone noticed the ingestion speed for v1.0.0 is slower than v0.8.3?
Several users have reported degradation in mutation speed. This is sometimes caused by inappropriate index use causing a large number of transactions to abort. If you tell a little bit about your data, ingestion pattern, schema etc, we may be able to suggest ways to improve ingestion speed.
FYI: When I load the tour data with
dgraph live -r 1million.rdf.gz --zero localhost:5080, dgraph v0.9.x took
50s to complete, but dgraph v1.0.0 took
1m33s to complete in the same laptop and OS.
we were writing to a store directly in v0.8.3, and as we begin to use transactions in 0.9.x and v1.x the ingesting speed drops down
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.