niveathika
commented
5 years ago @p-litvinov Was this run on SP editor or worker runtime?
Open p-litvinov opened 5 years ago
niveathika
commented
5 years ago @p-litvinov Was this run on SP editor or worker runtime?
p-litvinov
commented
5 years ago I had tried to run it on both runtimes: editor and worker. Every runtime has the issue in 4.4.0. But with 4.3.0 version it works without any exception with worker and editor.
niveathika
commented
5 years ago It seems the data source is not initialized properly, did you see any other errors? Was the app deployed successfully?
p-litvinov
commented
5 years ago Yes. App is deployed successfully and data source is initialized properly. I have another app where this aggregation with the same data source works in write mode. And it works fine. In the 4.3.0 both apps work fine with the data source.
p-litvinov
commented
5 years ago Maybe in the 4.4.0 it needs some additional rule in the app when using an aggregation in the read only mode?
niveathika
commented
5 years ago No, it does not, read-only mode need not connect to a source that's all. Siddi app with the read-only aggregation lets define an event table and see if we can connect to the data source.
This is not reproducible in our end.
Description: The error occured in WSO2-SP 4.4.0 when trying to use read-only aggregation from @store. In the WSO2-SP 4.3.0 it works fine.
Suggested Labels:
Suggested Assignees:
Affected Product Version:
OS, DB, other environment details and versions:
Steps to reproduce: Create a aggregation with store in DB Postgres or Mysql. Put several events in the aggregation. Run the second script with read-only aggregation: `@App:name("demoError1") @App:description("Aggregation with store exception.")
-- Trade aggregation, read-only --
define stream TradesVolumeFake( timestamp long, exchange string, client_id string, trade_volume double ); @store(type = 'rdbms' , datasource = 'AF2_DB') @purge(enable="false") define aggregation TradeClientAggregation from TradesVolumeFake select exchange, client_id, sum(trade_volume) as totalVolume, max(trade_volume) as maxVolume, min(trade_volume) as minVolume, count() as countTrades group by exchange, client_id aggregate by timestamp every sec ... year; define trigger TriggerPreLoadTrades at 'start'; define stream RebuildCmd (t int);
@sink(type='log', prefix='CalcStream') define stream CalcStream(_timestamp long, exchange string, client_id string, countTrades long); from TriggerPreLoadTrades select 1 as t insert into RebuildCmd; from RebuildCmd as cmd right outer join TradeClientAggregation as agg on not (agg.exchange is null) --agg.exchange =='moex.eq' and agg.client_id == 'BC15981' within -- '2019-05-09 00:00:00', '2019-05-15 00:00:00' -- for tests only -- time:timestampInMilliseconds(time:dateSub(time:dateFormat(currentTimeMillis(), 'yyyy-MM-dd 00:00:00.000'),100,'DAY','yyyy-MM-dd HH:mm:ss.SSS')), -- 100 DAYS is parameter time:timestampInMilliseconds(time:dateFormat(currentTimeMillis(), 'yyyy-MM-dd 00:00:00.000')) per 'minutes' select agg.AGG_TIMESTAMP as _timestamp, agg.exchange as exchange, agg.client_id as client_id, agg.countTrades as countTrades --- time:dateFormat(agg.AGG_TIMESTAMP, 'yyyy-MM-dd HH:mm:ss') as time order by _timestamp ---limit 10 insert into CalcStream;`
Related Issues: