google cloud datastore - Modeling an invite schema with embedded collections with dynamodb or docuemntdb -

i'm investigating whether use aws dynamodb or azure documentdb or google cloud price , simplicity app , wondering best approach typical invite schema.

an invite has

• userid : key (who created invite)
• gameid : key
• invitationlist : collection of userids

the queries running are

1. get invites userid == me
2. get invites userid in invitationlist

in mongo, set index on embedded invitationlist, , in sql set join table of gameid , invited userids.

using dynamodb or documentdb, in 1 "table" or have set second denormalized table 1 has invited userid per row set of invitedgameids?

e.g.

a secondary table with

• inviteduserid : key
• gameids : collection

similar hslriksen's answer, if criteria met, recommend denormalize of single document. criteria are:

1. the invitationlist games cannot grow unbounded.
2. even if it's bounded, maximum length array fit in document , transaction limits.

however, different hslriksen, recommend example document this:

{   gameid: <some game key>,   userid: <some user id>,   invitationlist: [<user id 1>, <user id 2>, ...] } 

you might decide use built-in id field games in case name above wrong.

the key difference between propose , hslriksen invitationslist pure array of foreign keys. allow indexes used array_contains clause in query.

note, in documentdb, tend store entity types in same big bucket , distinguish them string type field or better, is_my_type boolean field.