I use janusgraph server, version is 0.6.0,  backend storage is hbase.

I have a vertex that links 200000 vertex. But I query count vertex only return 1. g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').out().count()

bf41caab-f154-493d-a8d7-c393a9d9ff3b is my supsernode, And I create index for nodes. 

Hi Xin pacino,

Could you please also provide the result of the following query to be sure that the edges you are looking for are not in-edges:

g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').bothE().count()

If you want just the count, the bothE() step will perform better than the .both() step because the adjacencies are already stored with the central vertex (unless JanusGraph has a so-called strategy to optimize .both().count() into .bothE().count()).

Best wishes,    Marc

Op dinsdag 30 juni 2020 om 13:59:32 UTC+2 schreef Xin pacino:

I use janusgraph server, version is 0.6.0,  backend storage is hbase.

I have a vertex that links 200000 vertex. But I query count vertex only return 1. g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').out().count()

bf41caab-f154-493d-a8d7-c393a9d9ff3b is my supsernode, And I create index for nodes. 

Hi  HadoopMarc

my result can be know from the picture

Best wishes,    Xin 

在2020年7月1日星期三 UTC+8 下午11:03:03<HadoopMarc> 写道：

Hi Xin pacino,

Could you please also provide the result of the following query to be sure that the edges you are looking for are not in-edges:

g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').bothE().count()

If you want just the count, the bothE() step will perform better than the .both() step because the adjacencies are already stored with the central vertex (unless JanusGraph has a so-called strategy to optimize .both().count() into .bothE().count()).

Best wishes,    Marc

Op dinsdag 30 juni 2020 om 13:59:32 UTC+2 schreef Xin pacino:

I use janusgraph server, version is 0.6.0,  backend storage is hbase.

I have a vertex that links 200000 vertex. But I query count vertex only return 1. g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').out().count()

bf41caab-f154-493d-a8d7-c393a9d9ff3b is my supsernode, And I create index for nodes. 

The picture is loss?

result is "1"

在2020年7月2日星期四 UTC+8 上午10:16:49<Xin pacino> 写道：

Hi  HadoopMarc

my result can be know from the picture

Best wishes,    Xin 

在2020年7月1日星期三 UTC+8 下午11:03:03<HadoopMarc> 写道：

Hi Xin pacino,

Could you please also provide the result of the following query to be sure that the edges you are looking for are not in-edges:

g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').bothE().count()

If you want just the count, the bothE() step will perform better than the .both() step because the adjacencies are already stored with the central vertex (unless JanusGraph has a so-called strategy to optimize .both().count() into .bothE().count()).

Best wishes,    Marc

Op dinsdag 30 juni 2020 om 13:59:32 UTC+2 schreef Xin pacino:

I use janusgraph server, version is 0.6.0,  backend storage is hbase.

I have a vertex that links 200000 vertex. But I query count vertex only return 1. g.V().has('node','obj_id','bf41caab-f154-493d-a8d7-c393a9d9ff3b').out().count()

bf41caab-f154-493d-a8d7-c393a9d9ff3b is my supsernode, And I create index for nodes.