ran@cns.umist.ac.uk (Bob Nutter) (06/06/90)
Hi! I've just spent a day and a half trying to get one of our 4000's to boot off a 3000 (as usual). Every time I got the same error: (blah, blah, blah, bytes loaded...) 81270001 Boot error: can't get UID -the code means 'operation attempted on unopened stream' I thought this was due to the fact that I had changed some acl's rather foolishly in our authorised area (which is hard linked to the os on that node). I'd gone around making sure everything that needed setuid'ing, etc was done. I replaced whole trees with links to other nodes, all to no avail. I eventually tracked the error down to emanating from netboot (by searching through the binary), but got no further. I was so bemused by this stage I though it would be a good idea to finally put 10.2 on the node. Then someone who had just added a 4000 sr9.7.1 node to our network, which wasn't yet catalogued wanted it to print via our nodes. I happened to catalogue it on the same node I couldn't boot diskless off and HEY PRESTO! As if by magic, the diskless 4000 started booting, and has been ok since. Can anyone out there explain this? If I hadn't had to spend most of my time fixing acls set by choosing 'closed' this wouldn't have confused me and I wouldn't have wasted my time :-<) bob ------------------------------------------------------------------------------- bob nutter, computer officer | dept. of computation, UMIST | "John was a little crippled midget manchester m60 1qd UK | lesbian boy, but he stood ten foot tel:+44 61 200 3386 | tall with a KNIFE!" email:ran@cns.umist.ac.uk | -Butthole Surfers -------------------------------------------------------------------------------
rees@dabo.ifs.umich.edu (Jim Rees) (06/07/90)
In article <1990Jun6.092223.24369@cns.umist.ac.uk>, ran@cns.umist.ac.uk (Bob Nutter) writes: > I've just spent a day and a half trying to get one of our 4000's to > boot off a 3000 (as usual). Every time I got the same error: > > (blah, blah, blah, bytes loaded...) > 81270001 Boot error: can't get UID One thing that sometimes helps in these cases is to run netman in such a way that you can read its error messages. Kill it and just run it (/sys/net/netman) in a window, for example.
kdavis@bert.HYWLMSD.WA.COM (Karen Davis) (06/14/90)
I had this problem booting a DN2500 diskless from a DN3000. I fixed the link /sys/net/netman.rc to link to netman.com_sh and the problem went away. Directory "/sys/net": sys type current type uid length name file unstruct 514 diskless_list file nil 9140 netboot file coff 20660 netman file rec 5088 netman.bin_sh file rec 4398 netman.com_sh file unstruct 508 sample_diskless_list link netmain_srvr "//daune/etc/netmain_srvr" link netman.rc "netman.com_sh" Karen Davis Honeywell MSD (206) 345-3439
betsym@apollo.HP.COM (Elizabeth Minahan) (06/27/90)
kdavis@bert.UUCP (Karen Davis) writes: > I had this problem booting a DN2500 diskless > from a DN3000. I fixed the link /sys/net/netman.rc > to link to netman.com_sh and the problem went away. This error usually occurs when the guaranteed sys5.3 environment is not resident on the node you're booting off. /sys/net/netman.bin_sh uses certain commands from the /sys5.3/bin directory. If this isn't there it will never complete the 'node_data setup, and you will get this error. Every node that the os is installed on, including those where sys5.3 was not chosen as an available environment, should have a minimal set of sys5.3 commands to support installation scripts. Betsy Minahan Networking Support/Chelmsford Response Center