NSLS-II / metadatastore

DEPRECATED: Incorporated into https://github.com/NSLS-II/databroker
Other
2 stars 11 forks source link

Enh v0 insert #222

Closed tacaswell closed 8 years ago

tacaswell commented 8 years ago

I do not understand the warning failure only on py2.7.

Both versions go through identical code paths up to where the warning should be raised.

@danielballan do you know if there are changes between how warnings are suppressed between 2 and 3?

danielballan commented 8 years ago

Hmm. That's new to me.

danielballan commented 8 years ago

What happens if you run run_tests.py with -W?

codecov-io commented 8 years ago

Current coverage is 90.45%

Merging #222 into master will increase coverage by +27.36%

@@             master       #222   diff @@
==========================================
  Files             7          7          
  Lines           856        869    +13   
  Methods           0          0          
  Messages          0          0          
  Branches        141        144     +3   
==========================================
+ Hits            540        786   +246   
+ Misses          297         52   -245   
- Partials         19         31    +12   
  1. 3 files in metadatastore were modified. more
    • Misses -177
    • Partials +15
    • Hits +162

Powered by Codecov. Last updated by ff38847...d8eab60

danielballan commented 8 years ago

Very nice test pattern. I would like to review this with fresh eyes tomorrow.

danielballan commented 8 years ago

Thanks for doing this. Consider my one line comment and then merge at will.

ghost commented 8 years ago

@danielballan @tacaswell Does this conclude the indexing work?

tacaswell commented 8 years ago

@arkilic Hopefully, but we will know once we get it into the wild.

ghost commented 8 years ago

@tacaswell Cool, I will start aligning the mdc once we feel good about it. For those that do not have the context, I am not concerned because the indexing discussed here as it deals with query result cursors, not inserts. Nothing critical that would require rebuilding the db indexes or migration.

tacaswell commented 8 years ago

I would also add that these indecies are things that can (should?) be tuned on a per-beamline basis.

ghost commented 8 years ago

Yes, I came across a tool that monitors the most common queries and allows us to optimize the application accordingly. So far it looks pretty nifty but pricy give the amount of data we store in mongo ($2.50 per GB/month)