fix: replicate legacy behavior for using cache with queries #613
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Corrects the issue pointed out in googleapis#586, that we weren't doing the right thing for deleted entities. Also corrects an issue noticed while fixing that, where the cache wasn't being updated with entities from queries. Behavior should now match legacy in both these regards. Fixes googleapis#586
google-cla
bot
added
the
cla: yes
product-auto-label
bot
added
the
api: datastore
crwilcox
approved these changes
Mar 23, 2021
| @@ -379,6 +379,18 @@ def _next_batch(self): | |||
| for result_pb in response.batch.entity_results | |||
| ] | |||
|
|
|||
| if result_type == RESULT_TYPE_FULL: | |||
| # If we cached a delete, remove it from the result set. This may come cause | |||
| # some queries to return less than their limit even if there are more | |||
There was a problem hiding this comment.
I believe this is an underlying condition of Datastore as well. 👍
| Returns: | ||
| Any: The NDB entity for this result, if it is cached, otherwise | ||
| `_KEY_NOT_IN_CACHE`. May also return `None` if entity was deleted which | ||
| will cause `None` to be recorded in the cache. |
There was a problem hiding this comment.
was the None return true of legacy as well then? Seems a bit odd but likely not problematic.
There was a problem hiding this comment.
Yes, this is true of legacy as well.
crwilcox
added
the
automerge
gcf-merge-on-green
bot
removed
the
automerge
rwhogg
added a commit
to rwhogg/python-ndb
that referenced
this pull request
Feb 22, 2023
In certain circumstances, we were not respecting use_cache for queries, unlike legacy NDB, which is quite emphatic about supporting them. (See https://github.com/GoogleCloudPlatform/datastore-ndb-python/blob/59cb209ed95480025d26531fc91397575438d2fe/ndb/query.py#L186-L187) In googleapis#613 we tried to match legacy NDB behavior by updating the cache using the results of queries. We still do that, but now we respect use_cache, which was a valid keyword argument for Query.fetch() and friends, but was not passed down to the context cache when needed. As a result, the cache could mysteriously accumulate lots of memory usage and perhaps even cause you to hit memory limits, even if it was seemingly disabled and it didn't look like there were any objects holding references to your query results. This is a problem for certain batch-style workloads where you know you're only interested in processing a certain entity once. Fixes googleapis#752
rwhogg
added a commit
to rwhogg/python-ndb
that referenced
this pull request
Feb 22, 2023
In certain circumstances, we were not respecting use_cache for queries, unlike legacy NDB, which is quite emphatic about supporting them. (See https://github.com/GoogleCloudPlatform/datastore-ndb-python/blob/59cb209ed95480025d26531fc91397575438d2fe/ndb/query.py#L186-L187) In googleapis#613 we tried to match legacy NDB behavior by updating the cache using the results of queries. We still do that, but now we respect use_cache, which was a valid keyword argument for Query.fetch() and friends, but was not passed down to the context cache when needed. As a result, the cache could mysteriously accumulate lots of memory usage and perhaps even cause you to hit memory limits, even if it was seemingly disabled and it didn't look like there were any objects holding references to your query results. This is a problem for certain batch-style workloads where you know you're only interested in processing a certain entity once. Fixes googleapis#752
rwhogg
added a commit
to rwhogg/python-ndb
that referenced
this pull request
Feb 22, 2023
In certain circumstances, we were not respecting use_cache for queries, unlike legacy NDB, which is quite emphatic about supporting them. (See https://github.com/GoogleCloudPlatform/datastore-ndb-python/blob/59cb209ed95480025d26531fc91397575438d2fe/ndb/query.py#L186-L187) In googleapis#613 we tried to match legacy NDB behavior by updating the cache using the results of queries. We still do that, but now we respect use_cache, which was a valid keyword argument for Query.fetch() and friends, but was not passed down to the context cache when needed. As a result, the cache could mysteriously accumulate lots of memory usage and perhaps even cause you to hit memory limits, even if it was seemingly disabled and it didn't look like there were any objects holding references to your query results. This is a problem for certain batch-style workloads where you know you're only interested in processing a certain entity once. Fixes googleapis#752
rwhogg
added a commit
that referenced
this pull request
Feb 27, 2023
In certain circumstances, we were not respecting use_cache for queries, unlike legacy NDB, which is quite emphatic about supporting them. (See https://github.com/GoogleCloudPlatform/datastore-ndb-python/blob/59cb209ed95480025d26531fc91397575438d2fe/ndb/query.py#L186-L187) In #613 we tried to match legacy NDB behavior by updating the cache using the results of queries. We still do that, but now we respect use_cache, which was a valid keyword argument for Query.fetch() and friends, but was not passed down to the context cache when needed. As a result, the cache could mysteriously accumulate lots of memory usage and perhaps even cause you to hit memory limits, even if it was seemingly disabled and it didn't look like there were any objects holding references to your query results. This is a problem for certain batch-style workloads where you know you're only interested in processing a certain entity once. Fixes #752
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Corrects the issue pointed out in #586, that we weren't doing the right
thing for deleted entities. Also corrects an issue noticed while fixing
that, where the cache wasn't being updated with entities from queries.
Behavior should now match legacy in both these regards.
Fixes #586