| BML-S | Help Login

Last entries

    20250401.1035

  1. java - How does Jetty handle Threads and ThreadPools (note: diagram is incorrect) - [s]

    2. 20250328.1027

  2. java|log|stderr > What is the simplest way to configure `java.util.logging` - [s]

    $ rg java.util.logging.properties ./nexus-3.78.2-04/bin/nexus.vmoptions -Djava.util.logging.config.file=etc/spring/java.util.logging.properties $ cat ./nexus-3.78.2-04/etc/spring/java.util.logging.properties # Empty java.util.logging.properties to prevent the log to stderr, so that # all logs will be delegated to pax logging JUL handler only

    3. 20250326.0746

  3. java - How to run jcmd without the JDK? (portable) - [s]

    cd $JAVA_HOME_17 ./bin/jlink --add-modules jdk.jcmd --output ~/Desktop/custom-jdk

    4. 20250325.1159

  4. jupyter|pandas|ai|ollama > A Beginner’s Guide to Integrating LLMs - [s]

    5. 20250319.1146

  5. postgres > Benchmarking|performance PostgreSQL Batch Ingest (copy vs. insert .. unnest) - [s]

    Memory and disk write overheads: COPY is designed to have the least impact on your system, writing less data to disk and not polluting shared_buffers. This is actually a big consideration! In fact, both the COPY methods write 62 MB of WAL for the one million row test, while INSERT writes 109 MB. This ~1.7x rule seems to hold across any ingest size. Upserts: If you need conflict handling (INSERT...ON CONFLICT), COPY isn't an option, and you'll need to stick with INSERT (unless you just want to ignore errors and you're happy with text COPY, in which case PostgreSQL 17 has your back with ON_ERROR).

    6. 20250319.0855

  6. Why can't I name a folder or file 'CON' in Windows? - [s]

    7. 20250306.1558

  7. postgres > Using SSL with a PostgreSQL DB instance - AWS RDS - [s]

[s] public (7)