Skip to content

Comments

DM-36395: Support incremental redefinition of multi-snap visits#447

Merged
timj merged 11 commits intomainfrom
tickets/DM-36395
May 4, 2023
Merged

DM-36395: Support incremental redefinition of multi-snap visits#447
timj merged 11 commits intomainfrom
tickets/DM-36395

Conversation

@timj
Copy link
Member

@timj timj commented Mar 22, 2023

Checklist

  • ran Jenkins
  • added a release note for user-visible changes to doc/changes

@timj timj force-pushed the tickets/DM-36395 branch 5 times, most recently from a5df38b to 50f4b4a Compare March 27, 2023 23:04
@timj timj force-pushed the tickets/DM-36395 branch from 50f4b4a to 47f4235 Compare April 13, 2023 19:46
@timj timj requested a review from ktlim April 13, 2023 19:46
instrument=first.instrument,
)
)
records.difference_update(set(exposures_in_group))
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Seems kind of a shame to have this by group, then merge it all into one big list, then group again. But I guess that keeps the interfaces simple.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, it does, although I think there is a proposal to remove this subclass completely in #453 so it's not worth trying to make it more efficient.

@timj timj force-pushed the tickets/DM-36395 branch 2 times, most recently from da4a605 to b5f370b Compare May 3, 2023 20:17
timj added 8 commits May 3, 2023 13:25
We have to close the Pool after we are done with it.
Previously the multi-snap visit was only defined when the second
snap arrived. Now we can define the multi-snap visit even if
we only have one snap (so we get two visits defined). When the
second snap arrives there will be three visits and the multi-snap
visit will be updated if --update-records is enabled.
…ries

Implies update-records and so should be used when exposures for
multi-snap visits are being ingested incrementally whilst
continually updating the visit definition.
We had not been updating it as new options were added.
This is the true incremental mode where we ingest a new file
and immediately run visit definition.
@timj timj force-pushed the tickets/DM-36395 branch from b5f370b to 1c7d142 Compare May 3, 2023 20:25
timj added 3 commits May 3, 2023 13:35
Since we know that it's a single dataId we can make it more
efficiently with DataCoordinate rather than querying registry
for a single value.
Rather than asking for every exposure in the visit, first
work out which exposures we have and then determine the
missing sequence numbers and query for those.
@timj timj merged commit 41e5c8d into main May 4, 2023
@timj timj deleted the tickets/DM-36395 branch May 4, 2023 03:14
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants