Problem
Running a datastore query with or without FetchOptions.Builder.withLimit(100) takes the same execution time! Why is that? Isn't the limit method intended to reduce the time to retrieve results!?
Test setup
I am locally testing the execution time of some datastore queries with Google's App Engine. I am using the Google Cloud SDK Standard Environment with the App Engine SDK 1.9.59.
For the test, I created an example entity with 5 indexed properties and 5 unindexed properties. I filled the datastore with 50.000 entries of a test entity. I run the following method to retrieve 100 of this entities by utilizing the withLimit() method.
public List<Long> getTestIds() {
List<Long> ids = new ArrayList<>();
FetchOptions fetchOptions = FetchOptions.Builder.withLimit(100);
Query q = new Query("test_kind").setKeysOnly();
for (Entity entity : datastore.prepare(q).asIterable(fetchOptions)) {
ids.add(entity.getKey().getId());
}
return ids;
}
I measure the time before and after calling this method:
long start = System.currentTimeMillis();
int size = getTestIds().size();
long end = System.currentTimeMillis();
log.info("time: " + (end - start) + " results: " + size);
I log the execution time and the number of returned results.
Results
When I do not use the withLimit() FetchOptions for the query, I get the expected 50.000 results in about 1740 ms. Nothing surprising here.
If I run the code as displayed above and use withLimit(100) I get the expected 100 results. However, the query runs about the same 1740 ms!
I tested with different numbers of datastore entries and different limits. Every time the queries with or without withLimit(100) took the same time.
Question
Why is the query still fetching all entities? I am sure the query is not supposed to get all entities even though the limit is set to 100 right? What am I missing? Is there some datastore configuration for that? After testing and searching the web for 4 days I still can't find the problem.
FWIW, you shouldn't expect meaningful results from datastore performance tests performed locally, using either the development server or the datastore emulator - they're just emulators, they don't have the same performance (or even the 100% equivalent functionality) as the real datastore.
See for example Datastore fetch VS fetch(keys_only=True) then get_multi (including comments)
Related
UPDATED:
I do not have an entry in the index.yaml, expecting zig-zag to work fine on equalities (which it appears to in the datastore UI) and here is the actual query I'm making:
UploadBlock.query(
UploadBlock.time_slices == timeslice, # timeslice is each value in the repeated property
UploadBlock.company == block.company,
UploadBlock.aggregator_serial == block.aggregator_serial,
UploadBlock.instrument_serial == block.instrument_serial)\
.fetch(10, keys_only=True)
I've cross-posted this because I didn't read the initial support page first, so this is also on the google forum at https://groups.google.com/forum/#!topic/google-appengine/-7n8Y_tzCgM
I have a datastore query I make as an integral part of my system. It's an equality filter on four properties. When I make this query for one set of criteria, the query just hangs. It does not hang for this set of criteria in the datastore viewer, so I can see that there are only two results to return. I can fetch both of these entities by id without problem, but the query always hangs, both in remote shell and in production. The query is running fine for a multitude of different criteria, it appears to just be this one set. I have attempted replacing both entities so that the criteria still matches, but they have different ids and I get the same result.
For some more specific information, the four properties are defined as follows:
company = ndb.KeyProperty(indexed=True)
aggregator_serial = ndb.StringProperty(indexed=True)
instrument_serial = ndb.StringProperty(indexed=True)
time_slices = ndb.IntegerProperty(indexed=True, repeated=True)
I'm running gcloud versions
Google Cloud SDK 207.0.0
app-engine-java 1.9.64
app-engine-python 1.9.71
app-engine-python-extras 1.9.69
beta 2018.06.22
bq 2.0.34
core 2018.06.22
gcloud
gsutil 4.32
Any help on what I should try for next steps would be greatly appreciated.
Thanks in advance!
I currently have a an application running in the Google App Engine Standard Environment, which, among other things, contains a large database of weather data and a frontend endpoint that generates graph of this data. The database lives in Google Cloud Datastore, and the Python Flask application accesses it via the NDB library.
My issue is as follows: when I try to generate graphs for WeatherData spanning more than about a week (the data is stored for every 5 minutes), my application exceeds GAE's soft private memory limit and crashes. However, stored in each of my WeatherData entities are the relevant fields that I want to graph, in addition to a very large json string containing forecast data that I do not need for this graphing application. So, the part of the WeatherData entities that is causing my application to exceed the soft private memory limit is not even needed in this application.
My question is thus as follows: is there any way to query only certain properties in the entity, such as can be done for specific columns in a SQL-style query? Again, I don't need the entire forecast json string for graphing, only a few other fields stored in the entity. The other approach I tried to run was to only fetch a couple of entities out at a time and split the query into multiple API calls, but it ended up taking so long that the page would time out and I couldn't get it to work properly.
Below is my code for how it is currently implemented and breaking. Any input is much appreciated:
wDataCsv = 'Time,' + ','.join(wData.keys())
qry = WeatherData.time_ordered_query(ndb.Key('Location', loc),start=start_date,end=end_date)
for acct in qry.fetch():
d = [acct.time.strftime(date_string)]
for attr in wData.keys():
d.append(str(acct.dict_access(attr)))
wData[attr].append([acct.time.strftime(date_string),acct.dict_access(attr)])
wDataCsv += '\\n' + ','.join(d)
# Children Entity - log of a weather at parent location
class WeatherData(ndb.Model):
# model for data to save
...
# Function for querying data below a given ancestor between two optional
# times
#classmethod
def time_ordered_query(cls, ancestor_key, start=None, end=None):
return cls.query(cls.time>=start, cls.time<=end,ancestor=ancestor_key).order(-cls.time)
EDIT: I tried the iterative page fetching strategy described in the link from the answer below. My code was updated to the following:
wDataCsv = 'Time,' + ','.join(wData.keys())
qry = WeatherData.time_ordered_query(ndb.Key('Location', loc),start=start_date,end=end_date)
cursor = None
while True:
gc.collect()
fetched, next_cursor, more = qry.fetch_page(FETCHNUM, start_cursor=cursor)
if fetched:
for acct in fetched:
d = [acct.time.strftime(date_string)]
for attr in wData.keys():
d.append(str(acct.dict_access(attr)))
wData[attr].append([acct.time.strftime(date_string),acct.dict_access(attr)])
wDataCsv += '\\n' + ','.join(d)
if more and next_cursor:
cursor = next_cursor
else:
break
where FETCHNUM=500. In this case, I am still exceeding the soft private memory limit for queries of the same length as before, and the query takes much, much longer to run. I suspect the problem may be with Python's garbage collector not deleting the already used information that is re-referenced, but even when I include gc.collect() I see no improvement there.
EDIT:
Following the advice below, I fixed the problem using Projection Queries. Rather than have a separate projection for each custom query, I simply ran the same projection each time: namely querying all properties of the entity excluding the JSON string. While this is not ideal as it still pulls gratuitous information from the database each time, generating individual queries of each specific query is not scalable due to the exponential growth of necessary indices. For this application, as each additional property is negligible additional memory (aside form that json string), it works!
You can use projection queries to fetch only the properties of interest from each entity. Watch out for the limitations, though. And this still can't scale indefinitely.
You can split your queries across multiple requests (more scalable), but use bigger chunks, not just a couple (you can fetch 500 at a time) and cursors. Check out examples in How to delete all the entries from google datastore?
You can bump your instance class to one with more memory (if not done already).
You can prepare intermediate results (also in the datastore) from the big entities ahead of time and use these intermediate pre-computed values in the final stage.
Finally you could try to create and store just portions of the graphs and just stitch them together in the end (only if it comes down to that, I'm not sure how exactly it would be done, I imagine it wouldn't be trivial).
I have three models of the same structure(each string property stores ~20 characters):
class A(ndb.Model):
_use_cache = False
_use_memcache = False
p1 = ndb.KeyProperty(indexed=False)
p2 = ndb.IntegerProperty(default=0)
p3 = ndb.IntegerProperty(default=6)
p4 = ndb.StringProperty(indexed=True)
p5 = ndb.StringProperty(indexed=True)
p6 = ndb.StringProperty(indexed=True)
p7 = ndb.StringProperty(indexed=True)
p8 = ndb.StringProperty(indexed=True)
I want to create 200,000 entities for testing the performance at scale
I planned to feed the data in multiple runs of my app(i stopped my app from app launcher and started it again). I decided this approach because i noticed that the memory(as shown in my Task Manager) does not come down even after putting all the data. It actually increases gradually as the data is put in sets of 10000. But, on stopping the app, memory gets released. I suspect memory leak issue but am not sure
I gave a for loop with a total entity count of 17550. i gave ndb.put_multi() for sets of 1000 entities. I got 7051 entities as per count(A.query().count())
Next, i tried adding 12870 entities by the same method(in sets of 1000).
But i got a count of 7109
My computer is Lenovo T430 with 8 GB RAM(Windows 7 Enterprise); so resources should not be an issue and i do not run anything else apart from app engine launcher and chrome.. i am using the 1.9.1 version of gae python sdk..
Did anyone else face similar development server entity scaling issues.. What is maximum that you were able to achieve..
CONCERN:
Although GAE docs say that the query(filter conditions),fetch_page(count) response time is dependent only on the total size of the fetched matched entities, but i have seen response time degradation when the total (matched and not matched)entity count increased keeping the fetched count same..
Don't even bother trying to do this on the dev server.
It's meaningless in terms of what you experience on the dev server and can not in anyway be compared to how production will run. The only value (assuming you can get all 200,000 records stored in a timeframe you can handle) is application runs as expected.
Secondly doingquery.count() will not immediately give you the expected results on either dev server or production as you will encounter the effects of eventual consistency.
I'm flummoxed.
I noticed today that some data I thought should be present in my production appengine app wasn't showing up. I connected to the app via the remote console and ran the queries manually. Sure enough it looked like I only had 15 of the 101 rows I was expecting to see.
Then I went to my admin console at appengine.google.com and fired up the datastore viewer with the following query:
SELECT * FROM Assignment where game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb')
The result I see is the first page of 20 results. I page through those results, and am able to see all 101 entities. HOORAY! My data is still there. BUT why then can't I access it via the db api? (NOTE: I've already tried clearing memcache via the memcache viewer, even though this particularly query isn't manually memcached)
From the remote console:
> from google.appengine.ext.db import GqlQuery
> GqlQuery("SELECT * FROM Assignment WHERE game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb')").count()
15
The remote console agrees with the app itself, which only seems to be able to see 15 of the expected 101 rows.
What gives?
UPDATE:
I suspect this might be an indexing issue. If I issue get_by_key_name for one of the missing rows, it subsequently shows up in db api queries.
> GqlQuery("SELECT * FROM Assignment WHERE game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb')").count()
15
> entities.Assignment.get_by_key_name('201212-assignment-135.9')
<entities.Assignment object at 0xa11eb6c>
> GqlQuery("SELECT * FROM Assignment WHERE game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb')").count()
16
So should I (or can I) rebuild my indexes to remedy this problem?
UPDATE #2:
I attempted to build a perfect index for this query, and have just verified that even when the query does use the just-built index (via query.index_list()), the results are still only limited to a small subset of items available via the datastore viewer. Infuriatingly, it's actually a different subset than is available with the previous index (20 items vs 15 items). So now adding an additional filter term results in an additional 5 rows returned. So dumb.
All indexes claim to be "serving" so there shouldn't be any reason that the indexes are this far off.
UPDATE #3:
Sometimes, using my new index, I'll get the right answer:
> GqlQuery("SELECT * FROM Assignment WHERE game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb') and user = 'zee'").count()
101
However if I issue this query 10 times, it comes back with the 'bad' results about half the time:
> GqlQuery("SELECT * FROM Assignment WHERE game = KEY('Game', '201212-foo') and player = KEY('Player', 'player-mb') and user = 'zee'").count()
16
So maybe its an issue of a bad/behind bigtable replica that I'm hitting half the time, or something else completely opaque that we won't get an answer to (appengine status does list a service disruption today), but I have a feeling that this will be fixed on its own. Will update again if it does.
FINAL UPDATE:
As I suspected, when I woke up this morning my app (and manual queries) now see a consistent, correct view of the data. Would still love an answer as to why this happened, but until I get that I'm going to chalk it up to internal Google bigtable weirdness.
I filed this issue against appengine to see if I can get an answer from someone in the know.
For HRD applications, this is working as intended. App Engine High Replication Datastore (HRD) stores your data synchronously in multiple datacenters. However, the delay from the time a write is committed until it becomes visible in all datacenters means that queries across multiple entity groups (non-ancestor queries) can only guarantee eventually consistent results. [1]
In your specific case, the discrepancy between the results from your application and the Admin Console Datastore Viewer is just because they most likely are reading from different Datastore servers with different consistency.
If you require a consistent view of your data, I advise taking a closer look into the article "Structuring Data for Strong Consistency"
[1] https://developers.google.com/appengine/docs/java/datastore/structuring_for_strong_consistency
Is GQL easy to learn for someone who knows SQL? How is Django/Python? Does App Engine really make scaling easy? Is there any built-in protection against "GQL Injections"? And so on...
I'd love to hear the not-so-obvious ups and downs of using app engine.
Cheers!
My experience with google app engine has been great, and the 1000 result limit has been removed, here is a link to the release notes:
app-engine release notes
No more 1000 result limit - That's
right: with addition of Cursors and
the culmination of many smaller
Datastore stability and performance
improvements over the last few months,
we're now confident enough to remove
the maximum result limit altogether.
Whether you're doing a fetch,
iterating, or using a Cursor, there's
no limits on the number of results.
The most glaring and frustrating issue is the datastore api, which looks great and is very well thought out and easy to work with if you are used to SQL, but has a 1000 row limit across all query resultsets, and you can't access counts or offsets beyond that. I've run into weirder issues, with not actually being able to add or access data for a model once it goes beyond 1000 rows.
See the Stack Overflow discussion about the 1000 row limit
Aral Balkan wrote a really good summary of this and other problems
Having said that, app engine is a really great tool to have at ones disposal, and I really enjoy working with it. It's perfect for deploying micro web services (eg: json api's) to use in other apps.
GQL is extremely simple - it's a subset of the SQL 'SELECT' statement, nothing more. It's only a convenience layer over the top of the lower-level APIs, though, and all the parsing is done in Python.
Instead, I recommend using the Query API, which is procedural, requires no run-time parsing, and makes 'GQL injection' vulnerabilities totally impossible (though they are impossible in properly written GQL anyway). The Query API is very simple: Call .all() on a Model class, or call db.Query(modelname). The Query object has .filter(field_and_operator, value), .order(field_and_direction) and .ancestor(entity) methods, in addition to all the facilities GQL objects have (.get(), .fetch(), .count()), etc.) Each of the Query methods returns the Query object itself for convenience, so you can chain them:
results = MyModel.all().filter("foo =", 5).order("-bar").fetch(10)
Is equivalent to:
results = MyModel.gql("WHERE foo = 5 ORDER BY bar DESC LIMIT 10").fetch()
A major downside when working with AppEngine was the 1k query limit, which has been mentioned in the comments already. What I haven't seen mentioned though is the fact that there is a built-in sortable order, with which you can work around this issue.
From the appengine cookbook:
def deepFetch(queryGen,key=None,batchSize = 100):
"""Iterator that yields an entity in batches.
Args:
queryGen: should return a Query object
key: used to .filter() for __key__
batchSize: how many entities to retrieve in one datastore call
Retrieved from http://tinyurl.com/d887ll (AppEngine cookbook).
"""
from google.appengine.ext import db
# AppEngine will not fetch more than 1000 results
batchSize = min(batchSize,1000)
query = None
done = False
count = 0
if key:
key = db.Key(key)
while not done:
print count
query = queryGen()
if key:
query.filter("__key__ > ",key)
results = query.fetch(batchSize)
for result in results:
count += 1
yield result
if batchSize > len(results):
done = True
else:
key = results[-1].key()
The above code together with Remote API (see this article) allows you to retrieve as many entities as you need.
You can use the above code like this:
def allMyModel():
q = MyModel.all()
myModels = deepFetch(allMyModel)
At first I had the same experience as others who transitioned from SQL to GQL -- kind of weird to not be able to do JOINs, count more than 1000 rows, etc. Now that I've worked with it for a few months I absolutely love the app engine. I'm porting all of my old projects onto it.
I use it to host several high-traffic web applications (at peak time one of them gets 50k hits a minute.)
Google App Engine doesn't use an actual database, and apparently uses some sort of distributed hash map. This will lend itself to some different behaviors that people who are accustomed to SQL just aren't going to see at first. So for example getting a COUNT of items in regular SQL is expected to be a fast operation, but with GQL it's just not going to work the same way.
Here are some more issues:
http://blog.burnayev.com/2008/04/gql-limitations.html
In my personal experience, it's an adjustment, but the learning curve is fine.