CODEWITHSUNDEEP
google-cloud-platformgoogle-cloud-firestoregoogle-cloud-datastore
Murray
Murray

Reputation: 313

"no matching index found" when using Google Cloud Firestore via Datastore library

I have a Google Cloud Platform project with Datastore configured in Firestore mode, where the following (Python) code produces a google.api_core.exceptions.FailedPrecondition: 400 no matching index found. error.

from google.cloud import datastore

store = datastore.Client(project='my-project')
query = store.query(kind='my-kind')
query.add_filter('start_date', '<', 1500000000)
results = query.fetch()
for r in results:  # error thrown here
    print(r)

As Firestore automatically indexes single fields (and I have not excluded the start_timestamp field from this automatic indexing) I would expect this to work. What am I doing wrong?

(Note that this works perfectly on another project where Datastore is in Datastore rather than Firestore mode; the project I happen to need to use unfortunately already has Firestore in place. As Firestore is supposed to be backwards-compatible with Datastore, I had hoped this would just work.)

Upvotes: 0

Views: 1554

Answers (2)

Juancki
Juancki

Reputation: 1872

As Jim Morrison indicated, you need to use the Firestore library.

from google.cloud import firestore
db = firestore.Client()
# [START get_simple_query]
docs = db.collection(u'my-collection').where(u'start_date', u'<=', 1500000000).stream()

for doc in docs:
    print(u'{} => {}'.format(doc.id, doc.to_dict()))

Upvotes: 1

Jim Morrison
Jim Morrison

Reputation: 2887

Unfortunately, indexes are different between Firestore native & Firestore in Datastore mode. You'll need to use the matching client to get your queries to work. I suggest you use the google-cloud-firestore python libraries instead of the google-cloud-datastore libraries.

Upvotes: 0

Related Questions