Discover and read the best of Twitter Threads about #oracledb
Most recents (2)
trying to reproduce a system issue that presents itself in a complex ETL when run at full scale on a given system is tough.
but it’s a task i can do sometimes, and want to be able to do more often.
but it’s a task i can do sometimes, and want to be able to do more often.
it’s an important task bc attaching a debugger to production is risky. and extended events is of limited use when some system symptoms are known but query/worker/memory condition contributors aren’t.
the rules of the game: i almost never get to work with the production system itself. can supply some low-impact information gather/logging tools.
probably won’t be able to work with production data in nonprofit database, either.
probably won’t be able to work with production data in nonprofit database, either.
when do the stories people believe about a work of art, regardless of artist intent or history of the piece, become a part of the work of art?
i don’t know.
i don’t know.
working with a database engine, i’m an absolute stickler for “design intent” as revealed by documentation and historical development.
i have a particular view of RDBMS cpu scheduling and memory management: unless documented otherwise (in detail with boundaries), if an user query workload with no external calls makes the system unresponsive due to cpu or memory saturation, it’s a bug.
