On Thu, 12 Dec 2002, [iso-8859-2] Martin MOKREJ� wrote: > Well, I have recreated the filesystem, so it doesn't complain about small > inode size, if you mean that. But, there are no volumes created yet, so it > creaps that there no inodes to salvage.
ok > It seems, bosserver died, only fileserver, volserver and afsd stayed > overnight. so only bosserver is malloc()ing itself to death? have a debugger? run it under a debugger and see where it dies? > source tree to afs disk on another fileserver. It seems, when the cache > cannot be allocated as expect (as defined in cachinfo file), the mv(1) > commands cannot proceed. Is this a feature or bug? Could afsd skip using cache > in that case and just move the data anyway? it probably could, but it would involve another code path to basically read and make rpcs directly instead of pushing it through the cache. sort of ugly to deal with a situation which shouldn't be happening. > _______________________________________________ OpenAFS-devel mailing list [EMAIL PROTECTED] https://lists.openafs.org/mailman/listinfo/openafs-devel
