Re: FreeBSD on a MaxAttach?
I noticed that the SNAP Servers are far better in performance than the MaxAttach. WinNT/2000 embedded was a nice idea, but its so bloated I think they screwed it up a bit in its efficiency. BSDI v4.3 and FreeBSD kernels are more up to the task (I have BSDI whick rocks in its own right). The mini iso does wonders for me for custom solutions. I wonder if Solaris 9 can hold a few candles to BSDI v4.3 or FreeBSD v4.6. We need a few articles to compare these notes!!! ;o Ken _ Join the worlds largest e-mail service with MSN Hotmail. http://www.hotmail.com To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: FreeBSD on a MaxAttach?
Chris Dillon wrote: We have some Quantum SNAP Servers which are exactly the same thing as the older MaxAttach boxes except with bigger IDE drives, and they're still running the custom version of FreeBSD on them. They actually perform better than our much heftier Windows NT 4 servers. They even perform better than the newer MaxAttach boxes which are now running a form of Win2K and have much heftier hardware. Uh... the version of FreeBSD on the Quantum boxes is probably the same version of FreeBSD that was on the InterJets... *cough*. -- Terry To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: FreeBSD on a MaxAttach?
On Fri, 21 Jun 2002, Terry Lambert wrote: Uh... the version of FreeBSD on the Quantum boxes is probably the same version of FreeBSD that was on the InterJets... *cough*. 2.2.something? :-) Whatever version it is, I'm impressed with how well it works. The only problem I have with the Quantum SNAP boxes is the total lack of being able to script any kind of setting of file ownerships or ACLs. You have to set those entirely through the web interface, which is entirely unacceptable when you want to do it for 2000 user home directories. The NT command-line ACL tools don't work, which is how I script that kind of thing on NT servers, and I've tried in vain to write a PERL script that actually accessed and parsed the web interface and sent back the appropriate POSTs. It almost works, but I gave up for the time being. The only other option would be to write something to run in the JVM that is on them, and I can't find any API documentation on setting file ownership or ACLs, not to mention I don't know Java well enough to write such a thing in the first place. :-) -- Chris Dillon - cdillon(at)wolves.k12.mo.us - cdillon(at)inter-linc.net FreeBSD: The fastest and most stable server OS on the planet - Available for IA32 (Intel x86) and Alpha architectures - IA64, PowerPC, UltraSPARC, and ARM architectures under development - http://www.freebsd.org No trees were harmed in the composition of this message, although some electrons were mildly inconvenienced. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: FreeBSD on a MaxAttach?
Having had to make Lilo boot Linux on these boards I have some familiarity with
them. They don't have a standard BIOS, so they don't support the standard
routines that the newer bootloader expects (e.g. memory sizing). If you have
more questions feel free to follow up off list - I doubt the particulars of
these boards or CDS' custom/hacked version of FreeBSD 3.0 are of much interest
to those on the list. If need be I can contact the person who designed the
board - I know it is a bit of a kludge but he was working on a very limited time
frame. Also you'll find that the MAC address is stored on the boot ROM. Maxtor
has moved from FreeBSD to the Windows SAK so the newer boxes are likely to
have full BIOS support (they could not keep any of the CDS developers to
maintain the FreeBSD code base).
On Thu, 20 Jun 2002, Bruce A. Mah wrote:
Sorry to interrupt various flamewars with some actual technical
discussion... :-)
At ${REALJOB}, we've got a couple of Maxtor MaxAttach boxes we're trying
to play with. These are dedicated NFS/SMB servers. Physically they are
1U boxes with four 70GB IDE disks on them (wd0, wd1, wd2, wd3). They
have Pentium (P55C) processors, 128MB of RAM, an on-board fxp device, no
slots, and no removable media devices, all this on what appears to be a
semi-custom motherboard. We managed to find a serial console with the
help of a multimeter and an oscilloscope. The OS appears to be a
stripped-down FreeBSD 3.X...they have some kind of concatenated disk
driver that seems similar to ccd(4).
For various reasons, we're trying to figure out how to get a stock
FreeBSD 4-STABLE on them. We tossed in a scratch disk with 4.5-STABLE
as the primary master disk; the machine wouldn't even give a loader
prompt. We also tried booting with the existing wd0 and wd1, and our
disk on the secondary master; we could boot, but got a kernel panic
during an attempted boot to single-user mode...I suspect in the
concatenated disk driver trying to do some consistency checking.
I should mention that with all four of the original disks installed,
it functions properly, if slowly, as an NFS server. We're trying not
to wipe out the existing boot disk until we have at least a warm, furry
feeling that this is going to work. We haven't gotten that yet.
In theory we could put a populated obj/ tree on the existing disks and
use this to do a installkernel/installworld, but this commits us to a
course of action really early without an easy way to back out if
something goes wrong (see last paragraph).
Has anyone played around with one of these boxes?
Thanks in advance!
Bruce.
PS. It's crossed my mind that the staff time involved in making this
work could quickly exceed the cost of buying equivalent (maybe even
better) normal hardware. :-)
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-hackers in the body of the message
Re: FreeBSD on a MaxAttach?
On Thu, 20 Jun 2002, Bruce A. Mah wrote: PS. It's crossed my mind that the staff time involved in making this work could quickly exceed the cost of buying equivalent (maybe even better) normal hardware. :-) s/could/will/ If I were you I'd look at the 1U dual Xeon servers from SuperMicro. Onboard Gigabit Ethernet, DDR SDRAM, 4 internal 3.5 drive bays plus slimline CDROM and floppy, 64bit/133Mhz PCI-X slots. They're /nice/. You can find benchmarks on them here: http://www.vampire.vanderbilt.edu/benchmarks.php Brandon D. Valentine -- http://www.geekpunk.net [EMAIL PROTECTED] ++[++-][++-].[+-][+-]+.+++..++ +.+[++-]++.+++..+++.--..+. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
Re: FreeBSD on a MaxAttach?
Based on the amount of effort we had to put in, I have to agree that you're going to have to need a _lot_ of hardware for the software effort to pay off. -Kip On Thu, 20 Jun 2002, Brandon D. Valentine wrote: On Thu, 20 Jun 2002, Bruce A. Mah wrote: PS. It's crossed my mind that the staff time involved in making this work could quickly exceed the cost of buying equivalent (maybe even better) normal hardware. :-) s/could/will/ If I were you I'd look at the 1U dual Xeon servers from SuperMicro. Onboard Gigabit Ethernet, DDR SDRAM, 4 internal 3.5 drive bays plus slimline CDROM and floppy, 64bit/133Mhz PCI-X slots. They're /nice/. You can find benchmarks on them here: http://www.vampire.vanderbilt.edu/benchmarks.php Brandon D. Valentine -- http://www.geekpunk.net [EMAIL PROTECTED] ++[++-][++-].[+-][+-]+.+++..++ +.+[++-]++.+++..+++.--..+. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-hackers in the body of the message
