1 |
Thank you Martin for a most helpful response. |
2 |
|
3 |
I had not entered in /etc/devfsd.conf "LOOKUP ppp MODLOAD" but |
4 |
now realise this would be necessary from 'ppp for Linux" by Paul Mackerrsa. |
5 |
|
6 |
I did not cd / before umount /mnt/gentoo. Now noted. |
7 |
I am now reluctant to # reboot. |
8 |
I had not made a # umount /proc. Now noted. |
9 |
|
10 |
re PROCFS & TMPFS at boot have: |
11 |
Mounting /proc |
12 |
Mounting devfs |
13 |
Mounting tmpfs |
14 |
I think this is what you asked. |
15 |
|
16 |
I have not had repeating messages on this install. |
17 |
Your advice re info required is noted. |
18 |
|
19 |
RE backup of gentoo and devfs: |
20 |
|
21 |
I have written a script to backup the whole system. Get error messages: |
22 |
tar: /dev/log: socket ignored |
23 |
tar: /lib/dev-state/log: socket ignored |
24 |
tar: 3 tmp files: socket ignored |
25 |
tar: /proc/2.....to......14/exe: Cannot read link: No such file or directory |
26 |
tar: /proc/801/fd/4: Cannot stat: No such file or directory |
27 |
tar: Error exit delayed from previous errors |
28 |
|
29 |
Backed up 3.65GB to another partition.. |
30 |
|
31 |
Would it be safe / ok to umount /proc before and mount after the backup? |
32 |
|
33 |
Would devfs and /proc be ok after a restore? |
34 |
|
35 |
Many thanks, |
36 |
Kevin |
37 |
|
38 |
|
39 |
|
40 |
|
41 |
----- Original Message ----- |
42 |
From: "Martin Schlemmer" <azarah@g.o> |
43 |
To: "Gentoo-Dev" <gentoo-dev@g.o> |
44 |
Sent: Saturday, January 05, 2002 8:46 AM |
45 |
Subject: Re: [gentoo-dev] Third try with build -r14 |
46 |
|
47 |
On Fri, 2002-01-04 at 21:02, Kevin Gordon wrote: |
48 |
> Hi, |
49 |
> I am trying to understand where I am at fault installing build -r14. |
50 |
> (1) I firstly installed a kernel with "/dev/pts file system for Unix 98 |
51 |
> PTYs" enabled. In Daniel's IBM notes re Advanced File Systems he states to |
52 |
> disable this. Your install notes for build -r14 cover all items listed by |
53 |
> Daniel but fail to note this item should be disabled. Having this item |
54 |
> enabled and devfs active, would this be fatal? Could it cause the boot |
55 |
> process to fail (see previous emails re emerge update). |
56 |
> |
57 |
No, it should just not be mounted, as devfs handles UNIX 98 |
58 |
PTYs automatically. |
59 |
|
60 |
|
61 |
> (2) While installing ppp I used "# mknod /dev/ppp c 108 0". Is this ok in |
62 |
a |
63 |
> devfs system? Could it cause the boot process to fail ? |
64 |
> |
65 |
|
66 |
No, although you should not need to do this, as devfs should |
67 |
create it when 'ppp_generic.o' module is loaded. |
68 |
|
69 |
|
70 |
> (3) I have previously stopped part way through the build -r14 install |
71 |
> process and noticed during "# exit" "umount /mnt/gentoo/boot; umount |
72 |
> /mnt/gentoo; reboot" that the / parition did not un-mount. Daniel's |
73 |
article |
74 |
> suggests "umount -r /mnt/gentoo" to place / in a ro state before |
75 |
rebooting. |
76 |
> Could it cause the next boot process to fail ? |
77 |
> |
78 |
|
79 |
Did you try to 'cd /' and then unmount it ? It will not unmount if |
80 |
'pwd' is /mnt/gentoo. Also make sure that |
81 |
there is nothing mounted (like /mnt/gentoo/proc) in |
82 |
/mnt/gentoo. If it did not unmount clean, boot should not fail, but |
83 |
just fsck the root partition. |
84 |
|
85 |
|
86 |
Also check that you have PROCFS and TMPFS compiled in, else |
87 |
boot will fail. |
88 |
|
89 |
|
90 |
> (4) My most recent problem: (see email emerge update) |
91 |
> I get a loop repeating "Caching service dependencies" |
92 |
> Did a control-C |
93 |
> Get "rm: cannot unlink './pppd.tdb' : read-only file system" |
94 |
> Hangs on "Initializing random number generator" |
95 |
> One of my "emerge update" processed baselayout-1.6.9 and also glibc. Are |
96 |
any |
97 |
> current problems with glibc which could corrupt my system? |
98 |
> |
99 |
|
100 |
Strange one ('loop repeating "Caching service dependencies"'). |
101 |
This should not happen, except if it is a old baselayout, and |
102 |
some service had a 'need' depend of itself. Otherwise it |
103 |
is a bug we are not aware of. If you still get this problem, |
104 |
create a tarball of the contents of /etc/init.d/* , and |
105 |
mail it to: |
106 |
|
107 |
azarah@g.o |
108 |
|
109 |
Also some more details of the error messages on the boot |
110 |
failure would help, what services you have in your BOOT and |
111 |
DEFAULT runlevel (/etc/runlevels/boot/* and |
112 |
/etc/runlevels/default/* ), etc. |
113 |
|
114 |
|
115 |
Greetings, |
116 |
|
117 |
-- |
118 |
|
119 |
Martin Schlemmer |
120 |
Gentoo Linux Developer, Desktop Team Developer |
121 |
Cape Town, South Africa |