4 ms·
Not just that but holy shit some of those commercial RTOSes have major issues. I work in aerospace and we recently used one where the whole system would crash a
by shittyadmin 8y ago
Not just that but holy shit some of those commercial RTOSes have major issues. I work in aerospace and we recently used one where the whole system would crash after ~230 days of uptime.
At least with Linux you're getting a system that's been used so much that all major issues like that are ironed out. Nothing beats a few million testers.
- jtwaleson 8y ago248 days of uptime by any chance? edit: I saw the same on a fleet of thousands of JVMs which hung on 100% CPU after 248 days very consistently. Closest thing to an explanation I ever got was perhaps it is storing uptime in hundredths of a second (why not ms???) in signed 32 bit integers, see: https://ma.ttias.be/248-days/ https://ma.ttias.be/248-days/ In the end we solved it by restarting with a cronjob between 2am and 4am after 247 days...
- LinuxBender 8y agoOne thing to look at is the sum of Anonpages if you have THP enabled. That was enabled by default after CentOS6.2. The usage itself isn't an issue, but there is a known memory leak in THP and the fragmentation can get wedged after a couple hundred days based on usage characteristics of the server.
- blub 8y agoUbuntu Linux has millions of testers, your custom version has only been tested by you and your customers :)
- eeZah7Ux 8y agoUnless you patch the kernel very heavily, no.
- yitchelle 8y agoHmm, doubt that there are millions of folks testing Ubuntu controlling a car.
- NeutronStar 8y agoSure, but the linux tools used have been tested well before today.
- SEJeff 8y agoMillion? For linux perhaps billion :)