Because XFS is BETTER (Re: opinion on XFS)
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Mark Hahn hahn at physics.mcmaster.caFri May 10 07:33:27 PDT 2002
- Previous message: To crash or not to crash
- Next message: Because XFS is BETTER (Re: opinion on XFS)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
> That's really an advantage. Space utilization is in fact quite important > for contemporary applications on UNIX. Especially since we tend to have > so many small files such as sources, etc. nowadays. The savings can be > significant. odd how; I would state the exact opposite: space utilization is becoming pointless because disk is embarassingly cheap, and files are getting bigger. I also do not see any non-mistake large directories (where large means over say, 4k files.) at least not in the scientific computing arena - mail and news servers are naturally quite a lot different in how they use files. do me, if your batch system is generating tens of thousands of result files, you should probably consider making your batch infrastructure a little smarter - for instance, a perl master script to manage those jobs, and coalesce the data. > It's very easy to crash a node with a suitable code, so I shouldn't have to > re-install it or manually fsck it every time it fails to reboot after such a > crash... similarly, this statement makes no sense to me either: why are your nodes crashing? shouldn't you try fixing your nodes so they don't crash? it is VERY MUCH not the norm for linux nodes to crash, so you are clearly doing something wrong.
- Previous message: To crash or not to crash
- Next message: Because XFS is BETTER (Re: opinion on XFS)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
