hi
Howto virtualize physical SCO 5.0.6 on ESX4?
Here is a good manual:
Blog du DS45: HOWTO SCO P2V in the real life
BTW has someone already try to virtualize SCO 5.0.6?
According to this manual above, I try to create a backup file of entire disk inside the file system, but it doesn't work:
scounix# mount -f NFS 192.168.1.246:nfs /mnt
scounix# cd / ; find ./ -print cpio -ovfB > /mnt/backupfile.cpio &
4445
scounix#
What's wrong?
Take the & off the end of the line, and see what error messages if any you get.
Also with this command you might copy the contents of /mnt onto /mnt.
1 Like
I've done:
scounix# cd / ; find ./ -print cpio -ovfB > /mnt/backupfile.cpio
There will be a 5120KB file, very quickly created.
It cannot be a copy of the whole system:
scounix# dfspace
/ : Disk space: 3402,84 MB of 4000,00 MB available (85,07%).
/stand : Disk space: 14,15 MB of 24,41 MB available (57,97%).
/usr2 : Disk space: 53678,51 MB of 64423,23 MB available (83,32%).
Open the cpio file with vi and see what's in it.
1 Like
scounix# mount -f NFS 192.168.1.246:nfs /mnt
scounix# cd / ; find ./ -print -name ./mnt -prune | cpio -ovfB > /mnt/backupfile.cpio &
red: without /mnt/backupfile.cpio I hope find SCO can use find with -prune
sienna: pipe must be if cpio -o
1 Like
Thx a lot. This command, it seems to work well:
scounix# cd / ; find ./ -print -name ./mnt -prune | cpio -ovfB > /mnt/backupfile.cpio &
Before, I've done the following command and hope is doing this job as well:
scounix# cd / ; find . | cpio -ovfB -O /mnt/test.cpio &
or are there any differences?
And now I have 2 backup files and both have the same size 2097152KB.
BTW is this command correct, to recover on other machine, I mean on VM?
# mount -f NFS 192.168.1.246:nfs /mnt
# cpio �ivdmlu < /mnt/backupfile.cpio &
There is a 2gb limit on file sizes in 5.0.6
1 Like
sorry i don't saw:
You ought find directory that size about 1.5G
You try:
example:
cd /; du -ms *
If example /home 600M and /usr 1000M
you can do for filesystem /
scounix# mount -f NFS 192.168.1.246:nfs /mnt
scounix# cd / ; find ./ -print -name ./home -prune -name ./usr -prune -xdev| cpio -ovfB > /mnt/backupfile.cpio &
scounix# cd / ; find ./home -print -xdev | cpio -ovfB > /mnt/backupfile_home.cpio &
scounix# cd / ; find ./usr -print -xdev| cpio -ovfB > /mnt/backupfile_usr.cpio &
if you haven't free space you can do backup.....cpio by one in the order.
And you can on target system do:
cpio -ivdmlu < /mnt/backupfile.cpio &
cpio -ivdmlu < /mnt/backupfile_home.cpio &
cpio -ivdmlu < /mnt/backupfile_usr.cpio &
other filesystems you can transfer the method or start new SCO and transfer next filesystems by ...nfs...ftp...rcp....
you can see man find and man cpio at system SCO.
1 Like
Thx again, but my SCO 5.0.6 has a Problem with -m
scounix# cd /; du -ms *
usage: du [-afkrsuxV] [name ...]
/ and /usr2 are bigger than 2GB:
scounix# dfspace
15563 Signal 31 - core dumped
14824 Signal 31 - core dumped
/ : Disk space: 3403,25 MB of 4000,00 MB available (85,08%).
/stand : Disk space: 14,15 MB of 24,41 MB available (57,97%).
/usr2 : Disk space: 53678,50 MB of 64423,23 MB available (83,32%).
I used UNIX SCO 16 years ago.
At other unix option -m count in MB
you can use -k it's print count in KB
do "man du"
My proposion is for count how many place used in directory and all files in
Option "-name ./xxxxx -prune" means print name files(and directory) without (exclude) all in directory ./xxxxxx
I wants that you find directory example /usr 1000M /xxxx 500M
in your filesystem "/" without /usr /xxxxx is 1000M, you can archive it in file less then 2GB.
it's a problem with you system
file core is file with information about problems with programs.
when you move your server it's no nessesery to copy.
Please "rm core".
My proposition do:
cd / ; find ./ -name core -print -exec rm {} \;
change name directory if it nesesery.
if you don't undestand please send output;
cd /; du -ks *
PLEASE DO THE CODE.
change name directory if it nesesery.
scounix# mount -f NFS 192.168.1.246:nfs /mnt
scounix# cd / ; find ./ -print -name ./home -prune -name ./usr -prune -xdev| cpio -ovfB > /mnt/backupfile.cpio &
scounix# cd / ; find ./xxxx -print -xdev | cpio -ovfB > /mnt/backupfile_xxxx.cpio &
scounix# cd / ; find ./usr -print -xdev| cpio -ovfB > /mnt/backupfile_usr.cpio &
if you haven't free space you can do backup.....cpio by one in the order.
And you can on target system do:
cpio -ivdmlu < /mnt/backupfile.cpio &
cpio -ivdmlu < /mnt/backupfile_xxxx.cpio &
cpio -ivdmlu < /mnt/backupfile_usr.cpio &
other filesystems you can transfer the method or start new SCO and transfer next filesystems by ...nfs...ftp...rcp....
you can see man find and man cpio at system SCO.
1 Like
after SCO restart, I'm getting this:
scounix# dfspace
/ : Disk space: 3066,48 MB of 4000,00 MB available (76,66%).
/stand : Disk space: 14,15 MB of 24,41 MB available (57,97%).
/usr2 : Disk space: 53678,50 MB of 64423,23 MB available (83,32%).
Total Disk Space: 56759,14 MB of 68447,65 MB available (82,92%).
scounix# cd /; du -ks *
0
2 $[LOGFILE}
3946 EFS
4 Main.dt
2 Personal.dt
14414 addon
86 bin
250 core
2221 cpm
788 dev
847 etc
2 gesamtsi
84 ibin
2 install
294 lib
3 local
2 lost+found
4 mailfolders
11161764 mnt
381291 opt
0 pmd
2 root.jc
5 sbin
10 shlib
10072 stand
49 tcb
2 test
1140 tftpboot
36 tmp
6 trash.dt
0 unix
2 unix.nlsym
6 user.log
36168 usr
8947288 usr2
72489 var
139170 xaa
205664 xab
The problems is, I cannot backup /usr2, because cpio cannot deal with files > 2GB,
so I've tried to split output file to 1GB files using this command, but it won't work:
scounix# cd / ; find ./usr2 -depth -print -xdev | cpio -ovfB | gzip -c | split -b 1000m > /mnt/backupfile_usr2.gz_ &
There will be a file with 0KB created.
Hi,
in your system ./xxxx can be:
./opt ./xaa ./xab ..... I don't see 3066,48 MB at filesystem "/"
du normal show more then df.............. it isn't normal.
bye
you can read man find, cpio, gzip, split
1 Like
hi
I checked man pages, but still don't know howto solve it.
This command creates a file with 0KB:
scounix# cd / ; find ./usr2 -depth -print -xdev | cpio -ovfB | gzip -c | split -b 1000m > /mnt/backupfile_usr2.gz_
There must be a problem with split, because this command works well:
scounix# cd / ; find ./usr2 -depth -print -xdev | cpio -ovfB | gzip -c > /mnt/backup_usr2.gz_
man split
split -- split a file into pieces
Syntax
======
split [ -l line_count ] [ -a suffix_length ] [ file [ name ] ]
split [ -b n[k|m] ] [ -a suffix_length ] [ file [ name ] ]
split [ -line_count ] [ -a suffix_length ] [ file [ name ] ]
Description
===========
The split command reads file and writes it in as many n-line pieces as
necessary (default 1000), onto a set of output files. The name of the first
output file is name with aa appended, and so on lexicographically. If no
output name is given, x is default.
The following options are recognized:
-line_count
Specifies the number of lines in the resulting file.
-a suffix_length
Specifies that suffix_length letters are to be used to form the
suffix portion of the filenames of the split files.
-b n Specifies that the file is to be split into n byte pieces.
-b nk Specifies that the file is to be split into n kilobyte pieces.
-b nm Specifies that the file is to be split into n megabyte pieces.
If no input file is given, or if a dash (-) is given instead, the standard
input file is used.
Exit values
===========
split returns 0 on success, 1 on failure.
Examples
========
Split the file testfile (124 lines long) into fileaa, fileab, and fileac
(50, 50, and 24 lines long respectively):
split -l 50 testfile file
See also
========
bfs(C), csplit(C)
Standards conformance
=====================
split is conformant with:
ISO/IEC DIS 9945-2:1992, Information technology - Portable Operating System
Interface (POSIX) - Part 2: Shell and Utilities (IEEE Std 1003.2-1992);
AT&T SVID Issue 2;
X/Open CAE Specification, Commands and Utilities, Issue 4, 1992.
SCO OpenServer Release 5.0.6 -- 1 August 2000
See also split(TCL): man TCL split
/ The root file system is 4000-3403 = 600megabytes.
/usr2 is 64423-53768 = 11gigabytes.
You can copy all of the root file system and /stand into a single file easily enough.
Use
du -a /usr2 |sort -r -n >/tmp/du.srt
To list all the files on /usr2 in descending order. Directories will also be shown. Use this (/tmp/du.srt) to decide how to split the data up. You can also use compress (or gzip) if you have it to reduce the file size.
If you install cdrecord, you could probably get all of /usr2 on a single dual layer dvd.
1 Like
Thx a lot, but howto restore du.srt using du command?
du.srt is only a text file containing a list of the files and directories on the /usr2 file system sorted into descending order by size (size being 512 character blocks.)
If you analyze it, you should be able to compress enough files to get the usage below the size of a single dvd.
1 Like
Thx, I've done and getting this output:
I have done the following and it seems to work:
scounix# cd / ; find ./ -depth -print -xdev | cpio -ovfB | gzip -c > /mnt/backup.gz &
scounix# cd / ; find ./usr2 -depth -print -xdev | cpio -ovfB | gzip -c > /mnt/backup_usr2.gz &