I have a realtime upsert table which reads data from kinesis.There are 2 partitions or 2 kinesis shards.I updated config but after I restarted broker controller and server I still see 2 servers being hit when the accountId is in the query.
This was the original config.
SELECT currencyCode,activityId, activityType, distinctcount(sendIdContactId2) as 'sendId,contactId_count', sum(totalOrderNum) as totalOrderNum_sum, sumprecision(totalOrderAmt, 19) as totalOrderAmt_sum FROM events WHERE accountId = 1140607508363 AND flowId = 'd13c223f-b3fb-4fac-8689-77b89ab08f82' AND recordType = 'attribution' AND softDelete = 'null' group by currencyCode,activityType,activityId LIMIT 0, 1000 option(timeoutMs=20000)SELECT currencyCode,activityId, activityType, distinctcount(sendIdContactId2) as 'sendId,contactId_count', sum(totalOrderNum) as totalOrderNum_sum, sumprecision(totalOrderAmt, 19) as totalOrderAmt_sum FROM events WHERE accountId = 1140607508363 AND flowId = 'd13c223f-b3fb-4fac-8689-77b89ab08f82' AND recordType = 'attribution' AND softDelete = 'null' group by currencyCode,activityType,activityId LIMIT 0, 1000 option(timeoutMs=20000)
I still see 2 servers queried
![Screenshot 2024-01-22 at 3 42 40 PM](https://github.com/apache/pinot/assets/2093096/3d093b81-a698-472d-a574-20afcad6c44a)
Can you check if the upstream is partitioned properly? You may read the segment ZK metadata through the rest API and check if each segment only contain data from a single partition
Labels : Documentation, Troubleshooting
I have a realtime upsert table which reads data from kinesis.There are 2 partitions or 2 kinesis shards.I updated config but after I restarted broker controller and server I still see 2 servers being hit when the accountId is in the query. This was the original config.
I tried the below config suggested on the following slack channel by @walterddr https://apache-pinot.slack.com/archives/C011C9JHN7R/p1704225307389139
SELECT currencyCode,activityId, activityType, distinctcount(sendIdContactId2) as 'sendId,contactId_count', sum(totalOrderNum) as totalOrderNum_sum, sumprecision(totalOrderAmt, 19) as totalOrderAmt_sum FROM events WHERE accountId = 1140607508363 AND flowId = 'd13c223f-b3fb-4fac-8689-77b89ab08f82' AND recordType = 'attribution' AND softDelete = 'null' group by currencyCode,activityType,activityId LIMIT 0, 1000 option(timeoutMs=20000)SELECT currencyCode,activityId, activityType, distinctcount(sendIdContactId2) as 'sendId,contactId_count', sum(totalOrderNum) as totalOrderNum_sum, sumprecision(totalOrderAmt, 19) as totalOrderAmt_sum FROM events WHERE accountId = 1140607508363 AND flowId = 'd13c223f-b3fb-4fac-8689-77b89ab08f82' AND recordType = 'attribution' AND softDelete = 'null' group by currencyCode,activityType,activityId LIMIT 0, 1000 option(timeoutMs=20000)