RFR: 6936: Parser should track statistics during parsing

Marcus Hirt hirt at openjdk.java.net
Wed Feb 24 12:32:39 UTC 2021


On Tue, 23 Feb 2021 09:24:55 GMT, Jean-Philippe Bempel <github.com+4610701+jpbempel at openjdk.org> wrote:

> Add new `IParserStats` interface implemented by `EventCollection`.
> That way, `IItemCollection` is not touched, but you need to cast the
> instance to `IParserStats`.
> Support of old (v0.x) JFR format.
> During parsing, we collect number of events per type, the size, number
> of chunks and version of the file.
> We also track for version 1.x/2.x the number of skipped events.
> `RecordingPrinter` tool class is modified to support a summary verbosity mode
> and `-summary` option to print statistics per event type.

Do we need the number of events per type? Isn't that the same as a count aggregate with a type filter?

-------------

PR: https://git.openjdk.java.net/jmc/pull/219


More information about the jmc-dev mailing list