Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.debian.user > #182290 > unrolled thread
| Started by | Guy Marcenac <guy@posteurs.com> |
|---|---|
| First post | 2017-06-18 19:40 +0200 |
| Last post | 2017-06-20 18:30 +0200 |
| Articles | 4 — 3 participants |
Back to article view | Back to linux.debian.user
cifs boot time mounts fail after ugrading to Stretch Guy Marcenac <guy@posteurs.com> - 2017-06-18 19:40 +0200
Re: cifs boot time mounts fail after ugrading to Stretch guy@posteurs.com - 2017-06-20 10:10 +0200
Re: cifs boot time mounts fail after ugrading to Stretch Greg Wooledge <wooledg@eeg.ccf.org> - 2017-06-20 14:20 +0200
Re: cifs boot time mounts fail after ugrading to Stretch Guy Marcenac <guy@posteurs.com> - 2017-06-20 18:30 +0200
| From | Guy Marcenac <guy@posteurs.com> |
|---|---|
| Date | 2017-06-18 19:40 +0200 |
| Subject | cifs boot time mounts fail after ugrading to Stretch |
| Message-ID | <tTMl4-88J-9@gated-at.bofh.it> |
Hello, I just upgraded from jesie to stretch, and everything went fine. Except my network mounts from fstab fail at boot time. If I # mount -a from the prompt it works. I tried by adding one of the mounts as a systemd mount unit, and it fails too. Here are the entries from /etc/fstab: //192.168.0.21/video/cine /root/cine/i580mt cifs rw,credentials=/etc/cifspasswd,_netdev 0 0 //192.168.0.21/video/tv /root/tv/i580mt cifs rw,credentials=/etc/cifspasswd 0 0 My syslog: systemd[1]: Started OpenBSD Secure Shell server. kernel: [ 3.626226] FS-Cache: Netfs 'cifs' registered for caching kernel: [ 3.626273] Key type cifs.spnego registered kernel: [ 3.626277] Key type cifs.idmap registered kernel: [ 3.630942] CIFS VFS: Error connecting to socket. Aborting operation. kernel: [ 3.630956] CIFS VFS: Error connecting to socket. Aborting operation. kernel: [ 3.631234] CIFS VFS: Error connecting to socket. Aborting operation. kernel: [ 3.631987] CIFS VFS: cifs_mount failed w/return code = -101 kernel: [ 3.631999] CIFS VFS: cifs_mount failed w/return code = -101 systemd[1]: root-tv-i580mt.mount: Mount process exited, code=exited status=32 kernel: [ 3.633183] CIFS VFS: Error connecting to socket. Aborting operation. kernel: [ 3.633275] CIFS VFS: Error connecting to socket. Aborting operation. kernel: [ 3.633311] CIFS VFS: cifs_mount failed w/return code = -101 kernel: [ 3.633866] CIFS VFS: cifs_mount failed w/return code = -101 kernel: [ 3.634036] CIFS VFS: cifs_mount failed w/return code = -101 systemd[1]: Failed to mount /root/tv/i580mt. systemd[1]: Dependency failed for Remote File Systems. systemd[1]: remote-fs.target: Job remote-fs.target/start failed with result 'dependency'. systemd[1]: root-tv-i580mt.mount: Unit entered failed state. -- guy
[toc] | [next] | [standalone]
| From | guy@posteurs.com |
|---|---|
| Date | 2017-06-20 10:10 +0200 |
| Message-ID | <tUmoy-63S-5@gated-at.bofh.it> |
| In reply to | #182290 |
Le 2017-06-18 19:30, Guy Marcenac a écrit : > Hello, > > I just upgraded from jesie to stretch, and everything went fine. > Except my network mounts from fstab fail at boot time. > > If I > # mount -a > from the prompt it works. > > I tried by adding one of the mounts as a systemd mount unit, and it > fails too. > > Here are the entries from /etc/fstab: > //192.168.0.21/video/cine /root/cine/i580mt cifs > rw,credentials=/etc/cifspasswd,_netdev 0 0 > //192.168.0.21/video/tv /root/tv/i580mt cifs > rw,credentials=/etc/cifspasswd 0 0 > > My syslog: > systemd[1]: Started OpenBSD Secure Shell server. > kernel: [ 3.626226] FS-Cache: Netfs 'cifs' registered for caching > kernel: [ 3.626273] Key type cifs.spnego registered > kernel: [ 3.626277] Key type cifs.idmap registered > kernel: [ 3.630942] CIFS VFS: Error connecting to socket. Aborting > operation. > kernel: [ 3.630956] CIFS VFS: Error connecting to socket. Aborting > operation. > kernel: [ 3.631234] CIFS VFS: Error connecting to socket. Aborting > operation. > kernel: [ 3.631987] CIFS VFS: cifs_mount failed w/return code = -101 > kernel: [ 3.631999] CIFS VFS: cifs_mount failed w/return code = -101 > systemd[1]: root-tv-i580mt.mount: Mount process exited, code=exited > status=32 > kernel: [ 3.633183] CIFS VFS: Error connecting to socket. Aborting > operation. > kernel: [ 3.633275] CIFS VFS: Error connecting to socket. Aborting > operation. > kernel: [ 3.633311] CIFS VFS: cifs_mount failed w/return code = -101 > kernel: [ 3.633866] CIFS VFS: cifs_mount failed w/return code = -101 > kernel: [ 3.634036] CIFS VFS: cifs_mount failed w/return code = -101 > systemd[1]: Failed to mount /root/tv/i580mt. > systemd[1]: Dependency failed for Remote File Systems. > systemd[1]: remote-fs.target: Job remote-fs.target/start failed with > result 'dependency'. > systemd[1]: root-tv-i580mt.mount: Unit entered failed state. The problem is that systemd does not wait for DHCP addresses to be obtained. Being new to systemd I could not succeed in finding the right parameters and dependencies in order to have the mounts wait till we have an IP address. I tried remote-fs.target, network-online.target, and other hacks with no success. This machine being keyboardless, I did not dare migrate to systemd/networkd in order to try systemd-networkd-wait-online. As a workaround, I ended up setting the shares to automount, which works fine, but I would prefer to have them mounted at the very beginning..... Any idea? -- Guy
[toc] | [prev] | [next] | [standalone]
| From | Greg Wooledge <wooledg@eeg.ccf.org> |
|---|---|
| Date | 2017-06-20 14:20 +0200 |
| Message-ID | <tUqiu-8w4-19@gated-at.bofh.it> |
| In reply to | #182396 |
On Tue, Jun 20, 2017 at 10:00:14AM +0200, guy@posteurs.com wrote:
> The problem is that systemd does not wait for DHCP addresses to be
> obtained.
> Being new to systemd I could not succeed in finding the right parameters
> and dependencies in order to have the mounts wait till we have an IP
> address.
> I tried remote-fs.target, network-online.target, and other hacks with no
> success.
I can't help with CIFS mounts from fstab specifically, but I did learn
a thing or two about network-online.target in Debian due to a similar
issue with nfs-kernel-server.
network-online.target only cares about interfaces that are listed as
"auto" in /etc/network/interfaces. Not allow-hotplug. It has to be
auto.
Once you mark your interfaces as auto instead of allow-hotplug, then you
can create a dependency on network-online.target and have it actually
work. I did so for the nfs-kernel-server package, which uses a systemd
service named nfs-server with an alias of nfs-kernel-server... and yes,
it took me a while to figure *that* mess out too. Lesson #2: never try
to create a drop-in override for an *alias* name of a service. You must
determine the actual non-alias name first.
$ systemctl status nfs-server
* nfs-server.service - NFS server and services
Loaded: loaded (/lib/systemd/system/nfs-server.service; enabled; vendor prese
Drop-In: /etc/systemd/system/nfs-server.service.d
`-waitlonger.conf
Active: active (exited) since Mon 2017-06-19 08:22:36 EDT; 23h ago
Process: 522 ExecStart=/usr/sbin/rpc.nfsd $RPCNFSDARGS (code=exited, status=0/
Process: 519 ExecStartPre=/usr/sbin/exportfs -r (code=exited, status=0/SUCCESS
Main PID: 522 (code=exited, status=0/SUCCESS)
Tasks: 0 (limit: 4915)
CGroup: /system.slice/nfs-server.service
$ cat /etc/systemd/system/nfs-server.service.d/waitlonger.conf
[Unit]
Wants= network-online.target
After= network-online.target
See <https://bugs.debian.org/860264> for more details.
[toc] | [prev] | [next] | [standalone]
| From | Guy Marcenac <guy@posteurs.com> |
|---|---|
| Date | 2017-06-20 18:30 +0200 |
| Message-ID | <tUucq-2wu-29@gated-at.bofh.it> |
| In reply to | #182408 |
Wow Greg, youre awesome ! Le 20/06/2017 à 14:10, Greg Wooledge a écrit : > On Tue, Jun 20, 2017 at 10:00:14AM +0200, guy@posteurs.com wrote: >> The problem is that systemd does not wait for DHCP addresses to be >> obtained. >> Being new to systemd I could not succeed in finding the right parameters >> and dependencies in order to have the mounts wait till we have an IP >> address. >> I tried remote-fs.target, network-online.target, and other hacks with no >> success. > > I can't help with CIFS mounts from fstab specifically, but I did learn > a thing or two about network-online.target in Debian due to a similar > issue with nfs-kernel-server. > > network-online.target only cares about interfaces that are listed as > "auto" in /etc/network/interfaces. Not allow-hotplug. It has to be > auto. I would never have guessed this allow-hotplug thing. Thank you very much. > Once you mark your interfaces as auto instead of allow-hotplug, then you > can create a dependency on network-online.target and have it actually > work. I did so for the nfs-kernel-server package, which uses a systemd > service named nfs-server with an alias of nfs-kernel-server... and yes, > it took me a while to figure *that* mess out too. Lesson #2: never try > to create a drop-in override for an *alias* name of a service. You must > determine the actual non-alias name first. > Hope I will remember. From now on, it looks like we will have to become familiar with systemd. > See <https://bugs.debian.org/860264> for more details. > Dont you feel a bit alone there? And sorry for the direct post to your mb, -- guy
[toc] | [prev] | [standalone]
Back to top | Article view | linux.debian.user
csiph-web