da@cs.brown.edu (David Ascher) (04/08/91)
How many of the NewsAdmins [tm] out there are in charge of systems which keep NetNews around for "long" periods of time? Like, say, a month or more for a lot of the medium-flow newsgroups, and less for the talk.romance.chat and such? I have heard about the SysV problems with 64k inodes limit. Are there similar problems on Ultrix 4.1 (I know, it's not SysV)? Other problems? Thanks for any info -- == David Ascher -- Brown University, Providence RI 02912 == Internet: dascher@brownvm.Brown.EDU (Internet) == UUCP: uunet!brunix!da == Bitnet: dascher@brownvm
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (04/09/91)
In article <DA.91Apr8092631@igor.cs.brown.edu> da@cs.brown.edu (David Ascher) writes: | I have heard about the SysV problems | with 64k inodes limit. Are there similar problems on Ultrix 4.1 (I | know, it's not SysV)? Other problems? The problem is not related to the o/s, but rather the 16 bit vs. 32 bit inode number. Some PC based SysV versions, and many V7, SysIII, early BSD versions use 16 bit inode numbers. There are several solutions, one of which is using multiple filesystems for your news, such as putting comp in a separate f/s and then mounting it. Modify your startup so that you don't shoot yourself in the foot if the mount fails. My solution is to make the base directory (/usr/spool/news) read only owned by root, and then within the mounted filesystem make the directories which are mount points also R/O, owned by root. This prevents writing where you don't mean to if the mount fails. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) "Most of the VAX instructions are in microcode, but halt and no-op are in hardware for efficiency"