glasstiger

We could use the below SQL as a workaround for percentile_disc() for the time being: ``` with percentileRow as (select cast (0.95 * count() as LONG) as rownum from latencies),...

found a 15-minute gap in the log: ``` 2022-02-23T22:22:58.398440Z E i.q.c.l.t.LineTcpConnectionContext [23] could not parse measurement, code INVALID_FIELD_SEPARATOR at 166 line (may be mangled due to partial parsing) is WEATHER7,location=u...

i know you will hate this comment, but looking at the changes there are quite a few things like new objects, setters/getters, conditional checks... just wondering if it really optimises...

we should repoen https://github.com/questdb/questdb/issues/4144

another failure: ``` 2024-01-22T11:21:53.8826150Z random seeds: 1321827620612L, 1705922513870L 2024-01-22T11:21:53.8827000Z 2024-01-22T11:21:53.869721Z E i.q.t.TestListener ***** Test Failed *****io.questdb.test.cutlass.pgwire.PGJobContextTest.testExecuteAndCancelSqlCommands[WITH_WAL] duration_ms=72373 : 2024-01-22T11:21:53.8828090Z java.lang.RuntimeException: Failed on 2024-01-22T11:21:53.8828520Z ddl: create table tab (b boolean, ts...

happened again on CI windows-other build. [178.txt](https://github.com/questdb/questdb/files/13740110/178.txt)

happened again. full log: [148.txt.zip](https://github.com/questdb/questdb/files/15296675/148.txt.zip)

happened again. full log: [251.txt.zip](https://github.com/user-attachments/files/15974803/251.txt.zip)

this issue seems to be the same as https://github.com/questdb/questdb/issues/1923

happened again (windows-other). full log: [358.txt.zip](https://github.com/user-attachments/files/17571154/358.txt.zip)