-----Original Message----- From: Cedric Blancher [mailto:[email protected]]
> IMO you blame the wrong people. You can have the same kind of > problems with any Illumos-based distribution if you activate a > zone and let the machine just sit there for a week or so or have > a lot filesystem activity using mmap(). Either way the machines > will choke themselves to memory starvation. The only workaround > we found are regular reboots (every 24h), or limit the > ZFS ARC to an absolute minimum. I don't think you understand. My proxy tier does almost no reads from the file system. There is no content on the server. The ARC is irrelevant. ZFS isn't consuming the memory and if it did, it would give it up, given enough memory pressure. The proxy tier is simply a proxy to another system in the back end which also runs nginx. The backend systems, run nginx, Postgres, and some other utility scripts to push bits where they need to go. The backend systems do, as reported by zpool iostat, about 10k writes/second (up to about 50MB/s). There are way more back end systems than proxy tier systems and the backend runs just fine. Most will achieve a year of uptime in November. I have two radically different behaviors, on the same hardware platform, running essentially the same software stack where the major difference is the fire hose of network activity. In addition, to say that a zone needs to be rebooted every twenty-four hours is nuts. If this is what you are doing, then I fully do not understand your position. j.
_______________________________________________ OpenIndiana-discuss mailing list [email protected] http://openindiana.org/mailman/listinfo/openindiana-discuss
