During tuning long time queries, we observed this query
select distinct RES.*, DEF.KEY_ as PROC_DEF_KEY_, DEF.NAME_ as PROC_DEF_NAME_, DEF.VERSION_ as PROC_DEF_VERSION_, DEF.DEPLOYMENT_ID_ as DEPLOYMENT_ID_,
VAR.ID_ as VAR_ID_,
VAR.NAME_ as VAR_NAME_,
VAR.VAR_TYPE_ as VAR_TYPE_,
VAR.REV_ as VAR_REV_,
VAR.PROC_INST_ID_ as VAR_PROC_INST_ID_,
VAR.EXECUTION_ID_ as VAR_EXECUTION_ID_,
VAR.TASK_ID_ as VAR_TASK_ID_,
VAR.BYTEARRAY_ID_ as VAR_BYTEARRAY_ID_,
VAR.DOUBLE_ as VAR_DOUBLE_,
VAR.TEXT_ as VAR_TEXT_,
VAR.TEXT2_ as VAR_TEXT2_,
VAR.LAST_UPDATED_TIME_ as VAR_LAST_UPDATED_TIME_,
VAR.LONG_ as VAR_LONG_
from ACT_HI_PROCINST RES
left outer join ACT_RE_PROCDEF DEF on RES.PROC_DEF_ID_ = DEF.ID_
left outer join ACT_HI_VARINST VAR ON RES.PROC_INST_ID_ = VAR.EXECUTION_ID_ and VAR.TASK_ID_ is null
order by RES.START_TIME_ desc, VAR.LAST_UPDATED_TIME_ asc
LIMIT 20 OFFSET 0;
This query search process instance with process variables.
This query without distinct return same count of record like with it.
But time for query decrease for 6 times.
This query name is "selectHistoricProcessInstancesByQueryCriteria" from HistoricProcessInstance.xml.
I think, distinct should be removed from this query.
