lexyeevee

troublesome fox girl

hello i like to make video games and stuff and also have a good time on the computer. look @ my pinned for some of the video games and things. sometimes i am horny on @squishfox


You must log in to comment.

in reply to @lexyeevee's post:

For a long time, MongoDB's default configuration would have it acknowledge data as written immediately, but only flush it to disk every couple of seconds, or when a buffer was full. This made it bemchmark very well, but it was incredibly easy to end up with inconsistent data. If you actually set the durability so it was safe to use in a prodction environment, its performance took a nosedive.

At my first job, we had one microservice that used MongoDB, largely because one senior dev wanted to play with the hot new technology. When he eventually got bored & abandoned it, I was the one who had to to figure out what the hell was happening. "Transactions that complete just before the service reboots have a small chance to be only partially comitted," is not any easy bug to hunt down! I spent maybe a month and a half investigating, then when I finally figured out that I needed to to turn up the durability config, transaction throughput dropped to the point where the DB could no longer support the workload. We had to rewrite the service using SQL. I'm still mad!