Database Export Behaviour

[vnium]

I have a few questions regarding exporting a Dgraph database via the GraphQL endpoint.

• Is querying and mutating while exporting supported?
• If data is mutated while the export process is still running, will these changes be included in the export?
• Are multiple concurrent exports supported?
• Can an export process be cancelled? Will this also clean up export files?
• What is the impact of the chosen file format (rdf, json)? Which is file format is faster to export/import, which uses less space?

[MichelDiz]

@vnium

Can an export process be cancelled?

There’s no way to cancel as far as I know.

@vnium

Will this also clean up export files?

Nope.

@vnium

What is the impact of the chosen file format (rdf, json)?

In which sense? The only impact is the syntax difference.

@vnium

Which is file format is faster to export/import, which uses less space?

We haven’t checked this. But pretty sure RDF is the fastest one to export.

The other questions I invoke @dmai

Cheers.

[vnium]

@vnium

• Is querying and mutating while exporting supported?
• If data is mutated while the export process is still running, will these changes be included in the export?

@dmai any information to these two questions?

[dmai (Daniel Mai)]

Exports can run in a live cluster with queries and mutations running too. The export happens at a snapshot of the data, so any new mutations that are committed after an export request won’t be part of the export.