Date: Sat, 08 Jul 95 19:31:42 EDT From: evanc@synapse.net (Evan Champion) To: hackers@freebsd.org Subject: Filesystems Message-ID: <199507082331.TAA20064@sentinel.synapse.net>
next in thread | raw e-mail | index | archive | help
I have been thinking a bit about big disks and filesystems etc. We have some large disks (ie: the 9 GB variety). We really like them because you pack a lot of data in to a small amount of space, and with only 7 devices on a SCSI-2/FAST chain, that's important to us. The problem is that when our BSD/OS server goes down (no matter how good a system is, things happen; the most recent was a failed SIMM), it takes forever to go through the FS checks. Since this is a major server, we can't really afford the down time. I was thinking a little bit about our FreeBSD servers and how when they go down they don't seem to spend all that long in the disk checks. Now, maybe it is just that they also don't have these darned 9 GB disks on them, but it was just an observation that I made. Does FreeBSD do some kind of journalling that allows quick recoveries? Also, I noted there is a LFS filesystem. Not having the LFS documentation around, the name sort of implies that LFS is a journalled filesystem. I also note that it is brand new with 4.4BSD. Does the LFS work and how good is it at quick recoveries? We're thinking of using for a very large file server so if there is something we can do to help on recoveries it would be greatly appreciated by us and our users. Evan -- Evan Champion evanc@synapse.net * Visit our World Wide Web Server Director, Internet Systems * at <URL:http://www.synapse.net>; Synapse Internet *
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199507082331.TAA20064>