Uh oh!
There was an error while loading. Please reload this page.
Move stagingLocation and tempLocation to GcsOptions - #31
Conversation
peihe
commented
Mar 8, 2016
davorbonaci
commented
Mar 9, 2016
Hm... Can you argue more why this is the right change?
|
peihe
commented
Mar 9, 2016
The reason is BigQueryIO needs a tempLocation option across runners. Another option is to pull both of them to a new PipelineOption. |
davorbonaci
commented
Mar 9, 2016
Sounds like we might need a job-wide temporary location. I think this needs some discussion. |
dhalperi
commented
Mar 10, 2016
This one is tough -- the Would be nice to be able to reuse the one from DataflowPipelineOptions if such exists. |
f95b969 to
a78a806CompareDon't close data channels from environment sessions
…tion with Flink executor
apache#31 [euphoria-core] Implementation of accumulator API
…tion with Flink executor
* feat: add new data record read metrics Adds the following metrics: - Initial partition from created to scheduled state time (in ms) - Data record commit timestamp to read time (in ms) - Data record read time to emitted time (in ms) * feat: refine logging Adds the partition token to all possible logging. Add temporary metric of data record count as well. * chore: spotless apply
This will allow BigQueryOptions to extend GcsOptions, and BigQueryIO can access them through BigQueryOptions.