The right query to be done is
res = vns.post_query( 'FlowSeriesTable', start_time=str(generator_obj.flow_start_time), end_time=str(generator_obj.flow_end_time), select_fields=['sum(bytes)', 'sum(packets)', 'flow_count'], where_clause='vrouter=%s'% vrouter)
But with that too, for 10K tcp syn packets(20K flows) at 1K pps, analytics showed only about 13-15K flows
If i send 1K tcp syn packets(2K flows) at 100 pps, all 2K flows are seen in analytics
With Ashok's binary, http://nodei31:8085/Snh_AgentStatsReq? showed that all flows are exported,
but there some drops seen in sandesh client stats http://nodei31:8085/Snh_SandeshUVECacheReq?x=ModuleClientState
The right query to be done is
'FlowSerie sTable' ,
start_ time=str( generator_ obj.flow_ start_time) ,
end_ time=str( generator_ obj.flow_ end_time) ,
select_ fields= ['sum(bytes) ', 'sum(packets)', 'flow_count'],
where_ clause= 'vrouter= %s'% vrouter)
res = vns.post_query(
But with that too, for 10K tcp syn packets(20K flows) at 1K pps, analytics showed only about 13-15K flows nodei31: 8085/Snh_ AgentStatsReq? showed that all flows are exported, nodei31: 8085/Snh_ SandeshUVECache Req?x=ModuleCli entState
If i send 1K tcp syn packets(2K flows) at 100 pps, all 2K flows are seen in analytics
With Ashok's binary, http://
but there some drops seen in sandesh client stats http://