My ISP installed a new server for me with 2 200G hard drives. I asked him to partition each one into 2 100G slices and to do the disk labels for the 1st slice. He wasn't able to allocate the last 20G. I tried to do it remote, but sysinstall's disklabel wouldn't work on the first disk. So I asked him to reboot the server on the install CD and retry. No luck, it just wouldn't work. The BIOS access mode is "auto". I retried without sysinstall, using the comand "disklabel -e ad0s1". I created a new label manually, saved it, created a new file system, mounted it ... it all worked fine. I really don't think that the ISP did anything wrong. He installed several smaller servers, for me, running freebsd 4.* without any problems. I had the same problem on another server running FreeBSD 5.2.1 at another ISP. When I searched "The freebsd-questions Archives", I found someone who has had the same problem with freebsd 5.2.1 http://lists.freebsd.org/mailman/htdig/freebsd-questions/2004-June/049050.html here us what disklabel gives me. I Added the G label manually 8 partitions: # size offset fstype [fsize bsize bps/cpg] a: 2097152 8752464 4.2BSD 2048 16384 28552 b: 8228176 524288 swap c: 204796557 0 unused 0 0 # "raw" part, don't edit d: 10485760 10849616 4.2BSD 2048 16384 28552 e: 10485760 21335376 4.2BSD 2048 16384 28552 f: 161480704 31821136 4.2BSD 2048 16384 28552 g: 11494717 193301840 4.2BSD 2048 16384 28552 Fix: Use the command "disklabel" from the command prompt. How-To-Repeat: Reinstall the system. I can't do that because the server is used in production.
Responsible Changed From-To: freebsd-bugs->freebsd-sysinstall Over to maintainer(s)
I believe there have been commits to libdisk sanitizing what data is = snagged from BIOS. I am unable to duplicate this problem in 8.0.=
sysinstall has been replaced by bsdinstall in FreeBSD 9.x. Closing.