Summary

After upgrading dotnet driver from 3.5.2 to 3.6.0, it fails to evaluate some previously working aggregations for timeseries collection.

Please provide the version of the driver. If applicable, please provide the MongoDB server version and topology (standalone, replica set, or sharded cluster).

How to Reproduce

Model: 
{

"t" : ISODate("2023-03-01T00:00:00.000+0000"),

"m" : {

"i" : 1796183759,

"t" : "33333333-3333-3333-3333-333333333333"

{color:#dfdfdf}},

"_id" : ObjectId("698a351c81559bc74e2cb2c9"),

"p" : { "$numberDecimal" : "1.01" {color:#dfdfdf}},

"o" : ISODate("2026-02-09T19:27:22.590+0000")

{color:#dfdfdf}}
h4. t = Timestamp
o = DataTimestamp
p = Price
h4.
LINQ Query

GetCollection().AsQueryable()
    .Where(filter)
    .GroupBy(g => g.Timestamp, v => v)
    .Select(g => g.OrderBy(o => o.DataTimestamp).Last())
    .OrderBy(o => o.Timestamp)
    .GroupBy(g => g.Timestamp.Truncate(DateTimeUnit.Minute, 15), v => v)
    .Select(g => new AggregatedPriceSeriesItem
    {         Timestamp = g.First().Timestamp,         Price = g.Average(i => i.Price)     })
    .OrderBy(o => o.Timestamp);
h4. Additional Background
The difference is in how the Timestamp = g.First().Timestamp is generated

v3.5.2 working aggregation
_db.getCollection("...").aggregate([{ "$match" : { "m.t" : "33333333-3333-3333-3333-333333333333", "m.i" : 1796183759, "t" : { "$gte" : { "$date" : "2023-03-01T00:00:00Z" {color:#dfdfdf}}, "$lt" : { "$date" : "2023-03-02T00:00:00Z" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$group" : { "_id" : "$t", "_elements" : { "$push" : "$$ROOT" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$project" : { "_v" : { "$arrayElemAt" : [{ "$sortArray" : { "input" : "$_elements", "sortBy" : { "o" : 1 {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, -1] {color:#dfdfdf}}, "_id" : 0 {color:#dfdfdf}} {color:#dfdfdf}}, { "$sort" : { "_v.t" : 1 {color:#dfdfdf}} {color:#dfdfdf}}, { "$group" : { "_id" : { "$dateTrunc" : { "date" : "$_v.t", "unit" : "hour" {color:#dfdfdf}} {color:#dfdfdf}}, "__agg0" : { "$first" : "$_v.t" {color:#dfdfdf}}, "__agg1" : { "$avg" : "$_v.p" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$project" : { "t" : "$__agg0", "p" : "$__agg1", "_id" : 0 {color:#dfdfdf}} {color:#dfdfdf}}, { "$sort" : { "t" : 1 {color:#dfdfdf}} {color:#dfdfdf}}])_

v3.6 not-working aggregation (returns empty result)

db.getCollection("PriceSeries").aggregate([{ "$match" : { "m.t" : "33333333-3333-3333-3333-333333333333", "m.i" : -1506222593, "t" : { "$gte" : { "$date" : "2023-03-01T00:00:00Z" {color:#dfdfdf}}, "$lt" : { "$date" : "2023-03-02T00:00:00Z" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$group" : { "_id" : "$t", "_elements" : { "$push" : "$$ROOT" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$project" : { "_v" : { "$arrayElemAt" : [{ "$sortArray" : { "input" : "$_elements", "sortBy" : { "o" : 1 {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, -1] {color:#dfdfdf}}, "_id" : 0 {color:#dfdfdf}} {color:#dfdfdf}}, { "$sort" : { "_v.t" : 1 {color:#dfdfdf}} {color:#dfdfdf}}, { "$group" : { "_id" : { "$dateTrunc" : { "date" : "$_v.t", "unit" : "hour" {color:#dfdfdf}} {color:#dfdfdf}}, "__agg0" : { "$first" : "$t" {color:#dfdfdf}}, "__agg1" : { "$avg" : "$_v.p" {color:#dfdfdf}} {color:#dfdfdf}} {color:#dfdfdf}}, { "$project" : { "t" : "$__agg0", "p" : "$__agg1", "_id" : 0 {color:#dfdfdf}} {color:#dfdfdf}}, { "$sort" : { "t" : 1 {color:#dfdfdf}} {color:#dfdfdf}}])