#699 closed defect (fixed)
NFS restoration doesn't work with Mindi 2.1.5-r3113 & Mondo 3.0.3
Reported by: | Dan | Owned by: | Bruno Cornec |
---|---|---|---|
Priority: | high | Milestone: | 3.0.4 |
Component: | mindi | Version: | 3.0.3 |
Severity: | critical | Keywords: | nfs, restore, mindi |
Cc: |
Description
Hi,
ISO files are getting created just fine on the NFS target, but booting Mindi from the 1st ISO image and trying to restore from NFS hosted images doesn't work. Please see attached screenshot / log. Obviously automatic & interactive don't work.
sh-4.2# /sbin/start-netfs SIOCADDRT: File exists PING 10.1.0.240 (10.1.0.240) 56(84) bytes of data. 64 bytes from 10.1.0.240: icmp_req=1 ttl=64 time=0.496 ms 64 bytes from 10.1.0.240: icmp_req=2 ttl=64 time=0.424 ms 64 bytes from 10.1.0.240: icmp_req=3 ttl=64 time=0.358 ms --- 10.1.0.240 ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2002ms rtt min/avg/max/mdev = 0.358/0.426/0.496/0.056 ms /sbin/start-netfs: line 111: Starting rpcbind daemon…: command not found mount.nfs: Protocol not supported /tmp/isodir///dummyserver-1.iso: No such file or directory sh-4.2#
Ubuntu 12.04.2 LTS Kernel 3.5.0-23-generic, x86_64 arch.
Installed Mondo by following: https://help.ubuntu.com/community/MondoMindi
Attachments (1)
Change History (15)
by , 12 years ago
Attachment: | ScreenHunter_17 Jun. 12 14.25.jpg added |
---|
comment:1 by , 12 years ago
Summary: | NFS restoration doesn't work with Mindi 2.1.5-r3113 → NFS restoration doesn't work with Mindi 2.1.5-r3113 & Mondo 3.0.3 |
---|
comment:2 by , 12 years ago
Victor reported that Dan added the following libnss path to deplist.txt file, and that solved the problem :
/lib/x86_64-linux-gnu/libnss_compat-2.15.so /lib/x86_64-linux-gnu/libnss_compat.so.2 /lib/x86_64-linux-gnu/libnss_dns-2.15.so /lib/x86_64-linux-gnu/libnss_dns.so.2 /lib/x86_64-linux-gnu/libnss_files-2.15.so /lib/x86_64-linux-gnu/libnss_files.so.2 /lib/x86_64-linux-gnu/libnss_hesiod-2.15.so /lib/x86_64-linux-gnu/libnss_hesiod.so.2 /lib/x86_64-linux-gnu/libnss_nis-2.15.so /lib/x86_64-linux-gnu/libnss_nis.so.2 /lib/x86_64-linux-gnu/libnss_nisplus-2.15.so /lib/x86_64-linux-gnu/libnss_nisplus.so.2 /usr/lib/x86_64-linux-gnu/libnss_compat.so /usr/lib/x86_64-linux-gnu/libnss_dns.so /usr/lib/x86_64-linux-gnu/libnss_files.so /usr/lib/x86_64-linux-gnu/libnss_hesiod.so /usr/lib/x86_64-linux-gnu/libnss_nis.so /usr/lib/x86_64-linux-gnu/libnss_nisplus.so
comment:3 by , 12 years ago
Status: | new → assigned |
---|
Should be fixed with rev [3151]. Please report, test packages on their way for Ubuntu 12.04.
comment:4 by , 12 years ago
I think that the following should be added too.
Ubuntu 12.10 Quantal Quetzal 32bit # locate libnss_compat | xargs ls -lh -rw-r--r-- 1 root root 30K janv. 28 13:38 /lib/i386-linux-gnu/libnss_compat-2.15.so lrwxrwxrwx 1 root root 21 janv. 28 13:38 /lib/i386-linux-gnu/libnss_compat.so.2 -> libnss_compat-2.15.so lrwxrwxrwx 1 root root 38 janv. 28 13:39 /usr/lib/i386-linux-gnu/libnss_compat.so -> /lib/i386-linux-gnu/libnss_compat.so.2 -rw-r--r-- 1 root root 40K janv. 28 13:39 /usr/lib/i386-linux-gnu/xen/libnss_compat_pic.a # locate libnss_files | xargs ls -lh -rw-r--r-- 1 root root 46K janv. 28 13:38 /lib/i386-linux-gnu/libnss_files-2.15.so lrwxrwxrwx 1 root root 20 janv. 28 13:38 /lib/i386-linux-gnu/libnss_files.so.2 -> libnss_files-2.15.so lrwxrwxrwx 1 root root 37 janv. 28 13:39 /usr/lib/i386-linux-gnu/libnss_files.so -> /lib/i386-linux-gnu/libnss_files.so.2 -rw-r--r-- 1 root root 78K janv. 28 13:39 /usr/lib/i386-linux-gnu/xen/libnss_files_pic.a
comment:5 by , 12 years ago
Ok, I modified again the list, in fact reducing it to only keep what is really mandatory in rev [3160].
In fact as all deps and links are automatically included, I just need to add the higher level libs in this case /usr/lib/i386-linux-gnu/libnss_files.so and /usr/lib/i386-linux-gnu/libnss_compat.so
Do you want new package for 12.10/12.04 ?
comment:6 by , 12 years ago
Just a short note to say that my company also ran into this today, we lost a couple of days to it, unfortunately. I'd love new packages, if someone would be kind enough to push them to the ppa?
comment:7 by , 12 years ago
I would have built the packages myself, but it looks like it's quite complicated, and chances are that I'd just screw something up!
comment:8 by , 12 years ago
Lee, you don't have to make a new package.
I guess that you already installed the test version from ftp://ftp.mondorescue.org/test/
More informations about that test version here.
Then just add in your /etc/mindi/deplist.d/minimal.conf file:
/lib/x86_64-linux-gnu/libnss_compat.so.2 /lib/x86_64-linux-gnu/libnss_files.so.2 /usr/lib/x86_64-linux-gnu/libnss_compat.so /usr/lib/x86_64-linux-gnu/libnss_files.so
Or replace your /etc/mindi/deplist.d/minimal.conf file by this one that you can download.
comment:9 by , 12 years ago
follow-up: 12 comment:10 by , 12 years ago
In fact the latest test version of mindi should already include the right files in the deplist, so by downloading the Ubuntu packages from ftp://ftp.mondorescue.org/test/ubuntu you should be able to verify that this bug is indeed solved, and help me close it ;-)
follow-up: 13 comment:11 by , 12 years ago
Hey, so it took a whle to get all the infrastructure around PXE booting up in our environment, without trashing every other machine in the network!
A couple of things struck out:
1: it seems this works now, I installed the test packages, and rebuilt the ISO(s), all good 2: Upon restoration, there's still an "rpcbind: command not found" error, but it continues regardless, and gets fairly far, I ran into the problems:
a) it wasn't completley unattended (I might be missing a boot time keyword, for that, although I did specify -H) b) it complained because of disk arrangement differences, probably that's another issue (screenshot: http://cl.ly/image/1S1k36030N2i/Screen%20Shot%202013-07-03%20at%204.47.29%20PM.png ) - I was given a warning option that unfortunately I wasn't foresightful enough to screenshot, about hard disk sizes being different, I answered no, and got the answer there (I expected answering no would try and complete the autonuke, regardless of the disk size differences)
I'm not sure why the disk size warning cropped up, probably that's a different issue, but the machine was unchanged, and was being PXEbooted for autonuking just minutes after the ISO upload to the NFS server was completed.
comment:12 by , 12 years ago
Replying to bruno:
In fact the latest test version of mindi should already include the right files in the deplist, so by downloading the Ubuntu packages from ftp://ftp.mondorescue.org/test/ubuntu you should be able to verify that this bug is indeed solved, and help me close it ;-)
I'm kinda buried this week but I promise to test the new packages during the weekend ;) I will revert with my findings. Thank you !
comment:13 by , 12 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
Replying to LeeHambley:
1: it seems this works now, I installed the test packages, and rebuilt the ISO(s), all good
Ok, so I'll close this BR. Thanks for your report.
2: Upon restoration, there's still an "rpcbind: command not found" error, but it continues regardless, and gets fairly far, I ran into the problems:
Humm. As the binary is checked, then it could be a missing dyn. lib here. Could you give me the result of ldd /sbin/rpcbind and check whether these libs are on the ramdik or not (in the logs e.g.)
a) it wasn't completley unattended (I might be missing a boot time keyword, for that, although I did specify -H)
For PXE, use the RESTORE keyword on your boot line.
b) it complained because of disk arrangement differences, probably that's another issue (screenshot: http://cl.ly/image/1S1k36030N2i/Screen%20Shot%202013-07-03%20at%204.47.29%20PM.png ) - I was given a warning option that unfortunately I wasn't foresightful enough to screenshot, about hard disk sizes being different, I answered no, and got the answer there (I expected answering no would try and complete the autonuke, regardless of the disk size differences)
I'd also need the log to understand where the error is occuring precisely. In general in automatic mode (RESTORE) it works better in that case (even if I'm unsure why it"s behaving differently here !)
comment:14 by , 12 years ago
Just a short note to apologise, my employer decided we should roll our own, so I didn't have any time to pursue this anymore.
screenshot with the error