Jim Carter wrote: > Here's another nasty hang that I first blundered into last night, > which probably is unrelated but which might be helpful. The default > max size of tmpfs is half your physical memory. If the UML RAM > image is in tmpfs, and is configured larger than this (remember to > leave space for inodes and for other clients like KDE), and you make > it use the RAM, the machine sounds like it's swapping a bunch of > stuff, and UML pretty quickly gets a "kernel mode signal 7", SIGBUS, > which is logical when you think about it. So don't do that. > > If all your tmpfs's add up to more than your swap space plus > physical memory, and you fill them up, then the *host* will get a > kernel panic. So don't do that either. > > This came up when I was doing an online update of my UML's distro, > which used a lot of I/O buffers. When I limited the configured RAM > size to fit in tmpfs, the online update was finished with no > problems. Swapoff did not hang the machine. But I can't be sure > whether the UML swapped anything; I kind of doubt it.
Our setup is entirely the opposite - the physical machine has quite enough free memory, and there's plenty of free space on tmpfs. The UML instance, however has rather low memory (64M) and it is using swap space itself. The host is not swapping anything. But thanks for the heads up, we'll keep this in mind as well. -- Naked ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click _______________________________________________ User-mode-linux-user mailing list User-mode-linux-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
