> I have an aggregation that uses limit at the end to keep the results under
> the 16MB limit.  However, when I run it in a sharded environment, it
> fails.  The $limit comes after a $sort - and I guess that mongos has to get
> the entire sorted shard results back and interleave with the other shards
> in order to produce the full sorted list that gets limited.

> Is there anyway around this?  Can I output the results of aggregation to a
> temporary collection?

> Here is the error message:

> mongos> db.ga_page_path.aggregate(match, group, sort, limit);
> {
>     "errmsg" : "exception: sharded pipeline failed on shard megaset3: {
> errmsg: \"exception: aggregation result exceeds maximum document size
> (16MB)\", code: 16389, ok: 0.0 }",
>     "code" : 16390,
>     "ok" : 0
> }

> I have an aggregation that uses limit at the end to keep the results under
> the 16MB limit.  However, when I run it in a sharded environment, it
> fails.  The $limit comes after a $sort - and I guess that mongos has to get
> the entire sorted shard results back and interleave with the other shards
> in order to produce the full sorted list that gets limited.

> Is there anyway around this?  Can I output the results of aggregation to a
> temporary collection?

> Here is the error message:

> mongos> db.ga_page_path.aggregate(match, group, sort, limit);
> {
>     "errmsg" : "exception: sharded pipeline failed on shard megaset3: {
> errmsg: \"exception: aggregation result exceeds maximum document size
> (16MB)\", code: 16389, ok: 0.0 }",
>     "code" : 16390,
>     "ok" : 0
> }