Re: [DRBD-user] Unable to init a node

[Yannis Milios]

>
> Yes, I have three nodes :
> root@dmz-pve1:~ # drbdmanage list-nodes
> +---
> -+
> | Name | Pool Size | Pool Free |
> | State |
> |---
> -|
> | dmz-pve1 |   unknown |   unknown |
> |ok |
> | dmz-pve2 |   unknown |   unknown |
> |ok |
> | dmz-pve3 |   unknown |   unknown |
> |ok |
> +---
> -+
> But pool size is unknown, I have a VG named drbdpool
>

Check your /etc/ drbdmanaged.cfg file and make sure that the storage type
defined  there, is configured properly.
Is it thin or thick LVM? If it's thin LVM how did you create it ?


- The control volume is happy and connected to the other nodes?
>
> Yes
>   LV Path/dev/drbdpool/.drbdctrl_0
>   LV Name.drbdctrl_0
>   VG Namedrbdpool
>   LV Path/dev/drbdpool/.drbdctrl_1
>   LV Name.drbdctrl_1
>   VG Namedrbdpool
> root@dmz-pve1:~ #  drbdadm status
> .drbdctrl role:Primary
>   volume:0 disk:UpToDate
>   volume:1 disk:UpToDate
>   dmz-pve2 role:Secondary
> volume:0 peer-disk:UpToDate
> volume:1 peer-disk:UpToDate
>   dmz-pve3 role:Secondary
> volume:0 peer-disk:UpToDate
> volume:1 peer-disk:UpToDate
> root@dmz-pve1:~ #  drbd-overview 2>/dev/null
>  0:.drbdctrl/0  Connected(3*) Prim(dmz-pve1)/Seco(dmz-pve2,dmz-pve3)
> UpTo(dmz-pve1)/UpTo(dmz-pve2,dmz-pve3)
>  1:.drbdctrl/1  Connected(3*) Prim(dmz-pve1)/Seco(dmz-pve2,dmz-pve3)
> UpTo(dmz-pve1)/UpTo(dmz-pve3,dmz-pve2)
>
>
Yes, control volumes seem to be OK and drbd nodes too, but that's unrelated
to the storage which you have dedicated to DRBD (/etc/drbdmanaged.cfg). You
need to make sure that, that one is also OK.



>
> - You can create resources on the command line and assign them?
>
> I don't understand your question.
> I try to create virtual machine on one node of the proxmox cluster via the
> proxmox web interface.
> I need information on the pve storage.cfg
>

Before adding/removing VMs based on DRBD storage, you have to make sure
that DRBD storage works first. This means you need to check drbd pool
status, and use   drbdmanage add-resource, drbdmanage add-volume commands
in CLI to add/remove DRBD volumes. If that does not work, then proxmox gui
will not work as well.


Y
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Jdt]

Le 20/03/2018 à 10:07, Roland Kammerer a écrit :
> On Mon, Mar 19, 2018 at 07:12:29PM +0100, Jdt wrote:
>>
>> [SNIP}
> Please do not start debugging in the GUI, start at the lowest level.
>
> - You did add other nodes besides the initial node, didn't you?

Yes, I have three nodes :
root@dmz-pve1:~ # drbdmanage list-nodes
++
| Name | Pool Size | Pool Free
| | State |
||
| dmz-pve1 |   unknown |   unknown
| |    ok |
| dmz-pve2 |   unknown |   unknown
| |    ok |
| dmz-pve3 |   unknown |   unknown
| |    ok |
++
But pool size is unknown, I have a VG named drbdpool

> - The control volume is happy and connected to the other nodes?

Yes
  LV Path    /dev/drbdpool/.drbdctrl_0
  LV Name    .drbdctrl_0
  VG Name    drbdpool
  LV Path    /dev/drbdpool/.drbdctrl_1
  LV Name    .drbdctrl_1
  VG Name    drbdpool
root@dmz-pve1:~ #  drbdadm status
.drbdctrl role:Primary
  volume:0 disk:UpToDate
  volume:1 disk:UpToDate
  dmz-pve2 role:Secondary
    volume:0 peer-disk:UpToDate
    volume:1 peer-disk:UpToDate
  dmz-pve3 role:Secondary
    volume:0 peer-disk:UpToDate
    volume:1 peer-disk:UpToDate
root@dmz-pve1:~ #  drbd-overview 2>/dev/null
 0:.drbdctrl/0  Connected(3*) Prim(dmz-pve1)/Seco(dmz-pve2,dmz-pve3)
UpTo(dmz-pve1)/UpTo(dmz-pve2,dmz-pve3)
 1:.drbdctrl/1  Connected(3*) Prim(dmz-pve1)/Seco(dmz-pve2,dmz-pve3)
UpTo(dmz-pve1)/UpTo(dmz-pve3,dmz-pve2)

> - Drbdmanage is started on the other nodes?

Yes
root@dmz-pve1:~ # ps aux Op | grep drbd | grep -v grep
root  2657  0.0  0.0  0 0 ?    S<   16:38   0:00
[drbd-reissue]
root  2863  0.0  0.0  0 0 ?    S    16:39   0:00
[drbd_w_.drbdctr]
root  2865  0.0  0.0  0 0 ?    S<   16:39   0:00
[drbd0_submit]
root  2871  0.0  0.0  0 0 ?    S<   16:39   0:00
[drbd1_submit]
root  2926  0.0  0.0 270096 22184 ?    Sl   16:39   0:05
/usr/bin/python /usr/bin/dbus-drbdmanaged-service
root  2930  0.0  0.0   4348  1488 ?    S    16:39   0:00
drbdsetup events2 all
root  3032  0.0  0.0  0 0 ?    S    16:40   0:00
[drbd_s_.drbdctr]
root  3036  0.0  0.0  0 0 ?    S    16:40   0:00
[drbd_r_.drbdctr]
root  3231  0.0  0.0  0 0 ?    S    16:41   0:00
[drbd_a_.drbdctr]
root  3232  0.0  0.0  0 0 ?    S<   16:41   0:00
[drbd_as_.drbdct]
root  3416  0.0  0.0  0 0 ?    S    16:42   0:00
[drbd_s_.drbdctr]
root  3419  0.0  0.0  0 0 ?    S    16:42   0:00
[drbd_r_.drbdctr]
root  3555  0.0  0.0  0 0 ?    S    16:43   0:00
[drbd_a_.drbdctr]
root  3556  0.0  0.0  0 0 ?    S<   16:43   0:00
[drbd_as_.drbdct]
on the three nodes.

> - "drbdmanage role" on all nodes shows something meaningful?

Yes
root@dmz-pve1:~ # drbdmanage role
Operation completed successfully
leader

root@dmz-pve2:~ # drbdmanage role
Operation completed successfully
potential leader (CL: dmz-pve1)

root@dmz-pve3:~ # drbdmanage role
Operation completed successfully
potential leader (CL: dmz-pve1)


> - You can create resources on the command line and assign them?

I don't understand your question.
I try to create virtual machine on one node of the proxmox cluster via
the proxmox web interface.
I need information on the pve storage.cfg
I try to add :
drbd: drbdpool
    content images,rootdir
    redundancy 3

or

drbd: drbdstorage
    content images,rootdir
    redundancy 3

without success.
With the first solution, a logical volume is created on VG drbdpool on
only one node but nothing appear in /var/lib/drbd.d as in the previous
version
With the second solution, there is an error accessing volume

In proxmox interface, I don't see anyway to add DRBD resources in
Datacenter Storage.
Any help will be appreciated.

>
> Regards, rck
> ___
> drbd-user mailing list
> drbd-user@lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
Regards, jdt
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Roland Kammerer]

On Mon, Mar 19, 2018 at 07:12:29PM +0100, Jdt wrote:
> 
> 
> Le 13/03/2018 à 17:05, Roland Kammerer a écrit :
> > On Tue, Mar 13, 2018 at 04:57:34PM +0100, Jdt wrote:
> >> Le 13/03/2018 à 16:21, Roland Kammerer a écrit :
> >>
> >>> On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
>  Hi,
> 
>  We had a three nodes Proxmox cluster in V4 (jessie) with drbd from 
>  Proxmox.
>  I choose to reinitialize this cluster with Proxmox V5 (stretch) with
>  drbd from Linbit.
> 
>  I install Proxmox from the ISO image on the three nodes. Proxmox nodes
>  are not associated yet in a cluster.
>  I install drbdmanage-proxmox as indicated in
>  https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
>  My network interfaces are well configured for drbd.
>  On the first node I create the drbdpool volume group as indicated in
>  https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
> 
>  But it's failed in initialization :
>  root@dmz-pve1:~ # drbdmanage init 192.168.2.10
> 
>  You are going to initialize a new drbdmanage cluster.
>  CAUTION! Note that:
>    * Any previous drbdmanage cluster information may be removed
>    * Any remaining resources managed by a previous drbdmanage installation
>      that still exist on this system will no longer be managed by 
>  drbdmanage
> 
>  Confirm:
> 
>    yes/no: yes
>  Error: External command failed:
>  drbdsetup new-resource .drbdctrl 0
>  Command output:
>    (stdout)
>  Command exited with exit_code 20
> 
>  Initialization failed
> >>> Failing at that early stage usually means that the wrong module is
> >>> loaded. cat /proc/drbd
> >> root@dmz-pve1:~ # cat /proc/drbd
> >> version: 8.4.7 (api:1/proto:86-101)
> >> srcversion: 2DCC561E7F1E3D63526E90D
> > Then I was right. drbdmanage always was drbd9 only. There should be a
> > drbd-dkms module in our proxmox repo. Install that one and make sure it
> > is loaded.
> >
> Thanks. I add pve-headers and drbd-dkms and I can initialize the new
> drbdmanage cluster.
> It will be nice to include this in
> https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox chapter.
> But, I got a lot of messages on my syslog about PVE storage  about every
> ten seconds :
> > Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> > split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> > Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $free_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> > Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $total_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
> > Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> > split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> > Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $free_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> > Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $total_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
> > Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> > split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> > Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $free_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> > Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> > $total_space in multiplication (*) at
> > /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
> In proxmox interface I have no idea of free space on my drbdpool VG.
> Initialize the cluster create /dev/drbdpool/.drbdctrl_0 and
> /dev/drbdpool/.drbdctrl_1.
> For the moment no cluster are created for drbd and for proxmox.

Please do not start debugging in the GUI, start at the lowest level.

- You did add other nodes besides the initial node, didn't you?
- The control volume is happy and connected to the other nodes?
- Drbdmanage is started on the other nodes?
- "drbdmanage role" on all nodes shows something meaningful?
- You can create resources on the command line and assign them?

Regards, rck
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Shafeek Sumser]

Hi

Checked whether drbd9 module is loaded.  cat /proc/drbd.  It should say
drbd9.
I think you are still using the 8.4 version.

Shafeek





On Mon, Mar 19, 2018 at 10:12 PM, Jdt 
wrote:

>
>
> Le 13/03/2018 à 17:05, Roland Kammerer a écrit :
>
> On Tue, Mar 13, 2018 at 04:57:34PM +0100, Jdt wrote:
>
> Le 13/03/2018 à 16:21, Roland Kammerer a écrit :
>
>
> On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
>
> Hi,
>
> We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
> I choose to reinitialize this cluster with Proxmox V5 (stretch) with
> drbd from Linbit.
>
> I install Proxmox from the ISO image on the three nodes. Proxmox nodes
> are not associated yet in a cluster.
> I install drbdmanage-proxmox as indicated 
> inhttps://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
> My network interfaces are well configured for drbd.
> On the first node I create the drbdpool volume group as indicated 
> inhttps://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
>
> But it's failed in initialization :
> root@dmz-pve1:~ # drbdmanage init 192.168.2.10
>
> You are going to initialize a new drbdmanage cluster.
> CAUTION! Note that:
>   * Any previous drbdmanage cluster information may be removed
>   * Any remaining resources managed by a previous drbdmanage installation
> that still exist on this system will no longer be managed by drbdmanage
>
> Confirm:
>
>   yes/no: yes
> Error: External command failed:
> drbdsetup new-resource .drbdctrl 0
> Command output:
>   (stdout)
> Command exited with exit_code 20
>
> Initialization failed
>
> Failing at that early stage usually means that the wrong module is
> loaded. cat /proc/drbd
>
> root@dmz-pve1:~ # cat /proc/drbd
> version: 8.4.7 (api:1/proto:86-101)
> srcversion: 2DCC561E7F1E3D63526E90D
>
> Then I was right. drbdmanage always was drbd9 only. There should be a
> drbd-dkms module in our proxmox repo. Install that one and make sure it
> is loaded.
>
>
> Thanks. I add pve-headers and drbd-dkms and I can initialize the new
> drbdmanage cluster.
> It will be nice to include this in https://docs.linbit.com/docs/
> users-guide-9.0/#ch-proxmox chapter.
> But, I got a lot of messages on my syslog about PVE storage  about every
> ten seconds :
>
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 220.
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 221.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 220.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 221.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 220.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at 
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm
> line 221.
>
> In proxmox interface I have no idea of free space on my drbdpool VG.
> Initialize the cluster create /dev/drbdpool/.drbdctrl_0 and
> /dev/drbdpool/.drbdctrl_1.
> For the moment no cluster are created for drbd and for proxmox.
>
> /etc/drbdmanaged.cfg contain :
>  [LOCAL]
> storage-plugin = drbdmanage.storage.lvm_thinlv.LvmThinLv
> force=1
>
> and /etc/pve/storage.cfg :
> dir: local
> path /var/lib/vz
> content iso,vztmpl,backup
>
> drbd: drbdstorage
> content images,rootdir
> redundancy 3
>
> Any Ideas ?
> Thanks in advance
>
> Regards, JdT
>
>
>
>
>
>
>
> ___
> drbd-user mailing list
> drbd-user@lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
>
>


-- 
Shafeek SUMSER
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Jdt]

Le 13/03/2018 à 17:05, Roland Kammerer a écrit :
> On Tue, Mar 13, 2018 at 04:57:34PM +0100, Jdt wrote:
>> Le 13/03/2018 à 16:21, Roland Kammerer a écrit :
>>
>>> On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
 Hi,

 We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
 I choose to reinitialize this cluster with Proxmox V5 (stretch) with
 drbd from Linbit.

 I install Proxmox from the ISO image on the three nodes. Proxmox nodes
 are not associated yet in a cluster.
 I install drbdmanage-proxmox as indicated in
 https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
 My network interfaces are well configured for drbd.
 On the first node I create the drbdpool volume group as indicated in
 https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster

 But it's failed in initialization :
 root@dmz-pve1:~ # drbdmanage init 192.168.2.10

 You are going to initialize a new drbdmanage cluster.
 CAUTION! Note that:
   * Any previous drbdmanage cluster information may be removed
   * Any remaining resources managed by a previous drbdmanage installation
     that still exist on this system will no longer be managed by drbdmanage

 Confirm:

   yes/no: yes
 Error: External command failed:
 drbdsetup new-resource .drbdctrl 0
 Command output:
   (stdout)
 Command exited with exit_code 20

 Initialization failed
>>> Failing at that early stage usually means that the wrong module is
>>> loaded. cat /proc/drbd
>> root@dmz-pve1:~ # cat /proc/drbd
>> version: 8.4.7 (api:1/proto:86-101)
>> srcversion: 2DCC561E7F1E3D63526E90D
> Then I was right. drbdmanage always was drbd9 only. There should be a
> drbd-dkms module in our proxmox repo. Install that one and make sure it
> is loaded.
>
Thanks. I add pve-headers and drbd-dkms and I can initialize the new
drbdmanage cluster.
It will be nice to include this in
https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox chapter.
But, I got a lot of messages on my syslog about PVE storage  about every
ten seconds :
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> Mar 19 18:54:02 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> Mar 19 18:54:12 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value in
> split at /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 217.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $free_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 220.
> Mar 19 18:54:22 dmz-pve1 pvestatd[1984]: Use of uninitialized value
> $total_space in multiplication (*) at
> /usr/share/perl5/PVE/Storage/Custom/DRBDPlugin.pm line 221.
In proxmox interface I have no idea of free space on my drbdpool VG.
Initialize the cluster create /dev/drbdpool/.drbdctrl_0 and
/dev/drbdpool/.drbdctrl_1.
For the moment no cluster are created for drbd and for proxmox.

/etc/drbdmanaged.cfg contain :
 [LOCAL]
storage-plugin = drbdmanage.storage.lvm_thinlv.LvmThinLv
force=1

and /etc/pve/storage.cfg :
dir: local
    path /var/lib/vz
    content iso,vztmpl,backup

drbd: drbdstorage
    content images,rootdir
    redundancy 3

Any Ideas ?
Thanks in advance

Regards, JdT






___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Yannis Milios]

>>  root@dmz-pve1:~ # modinfo drbd
>> filename:   /lib/modules/4.13.13-6-pve/kernel/drivers/block/drbd/
drbd.ko
>> alias:  block-major-147-*
>> license:GPL
>> version:8.4.7   < (this is wrong!)


Since version 5, PVE ships with drbd8 kernel module (see version section
above) . That's the reason you're getting errors during drbd cluster
initialization.
You will have to install drbd-dkms package from Linbit's repository in
order to get drbd9 kernel module installed. Make sure you install
pve-headers-x package before doing that, otherwise drbd9 kernel module
will fail to build.

Y
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Jean-Daniel TISSOT]

And i forget, I get the same result


Le 13/03/2018 à 16:57, Jdt a écrit :
>
> Le 13/03/2018 à 16:21, Roland Kammerer a écrit :
>
>> On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
>>> Hi,
>>>
>>> We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
>>> I choose to reinitialize this cluster with Proxmox V5 (stretch) with
>>> drbd from Linbit.
>>>
>>> I install Proxmox from the ISO image on the three nodes. Proxmox nodes
>>> are not associated yet in a cluster.
>>> I install drbdmanage-proxmox as indicated in
>>> https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
>>> My network interfaces are well configured for drbd.
>>> On the first node I create the drbdpool volume group as indicated in
>>> https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
>>>
>>> But it's failed in initialization :
>>> root@dmz-pve1:~ # drbdmanage init 192.168.2.10
>>>
>>> You are going to initialize a new drbdmanage cluster.
>>> CAUTION! Note that:
>>>   * Any previous drbdmanage cluster information may be removed
>>>   * Any remaining resources managed by a previous drbdmanage installation
>>>     that still exist on this system will no longer be managed by drbdmanage
>>>
>>> Confirm:
>>>
>>>   yes/no: yes
>>> Error: External command failed:
>>> drbdsetup new-resource .drbdctrl 0
>>> Command output:
>>>   (stdout)
>>> Command exited with exit_code 20
>>>
>>> Initialization failed
>> Failing at that early stage usually means that the wrong module is
>> loaded. cat /proc/drbd
> root@dmz-pve1:~ # cat /proc/drbd
> version: 8.4.7 (api:1/proto:86-101)
> srcversion: 2DCC561E7F1E3D63526E90D
>
>>> In /etc/drbdmanaged.cfg, I have
>> You should not.
> Ok, I reinstall
>
> root@dmz-pve1:~ # modinfo drbd
> filename:  
> /lib/modules/4.13.13-6-pve/kernel/drivers/block/drbd/drbd.ko
> alias:  block-major-147-*
> license:    GPL
> version:    8.4.7
> description:    drbd - Distributed Replicated Block Device v8.4.7
> author: Philipp Reisner , Lars Ellenberg
> 
> srcversion: 2DCC561E7F1E3D63526E90D
> depends:    lru_cache,libcrc32c
> retpoline:  Y
> intree: Y
> name:   drbd
> vermagic:   4.13.13-6-pve SMP mod_unload modversions
> parm:   minor_count:Approximate number of drbd devices (1-255)
> (uint)
> parm:   disable_sendpage:bool
> parm:   allow_oos:DONT USE! (bool)
> parm:   proc_details:int
> parm:   usermode_helper:string
>
> root@dmz-pve1:~ # cat /etc/drbdmanaged.cfg
> [LOCAL]
> storage-plugin = drbdmanage.storage.lvm_thinlv.LvmThinLv
> force=1
> root@dmz-pve1:~ # lsmod | grep drbd
> drbd  360448  0
> lru_cache  16384  1 drbd
> libcrc32c  16384  2 dm_persistent_data,drbd
>
>
>
>>
>> If you don't have "force" set, most of them are ignored. Having options
>> per node, per config files isn't possible for years. As I don't see
>> "force", but your "custom options", that indicates to me that you did
>> not use our version of drbdmanage for proxmox, which sets up
>> drbdmanaged.cfg correctly, or you just overwrote what was in it with
>> what you thought is a good idea. Unfortunately, it was not a good
>> idea...
>> Regards, rck
>> ___
>> drbd-user mailing list
>> drbd-user@lists.linbit.com
>> http://lists.linbit.com/mailman/listinfo/drbd-user
> Regards, JdT
>
>
> ___
> drbd-user mailing list
> drbd-user@lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user

-- 
Bien cordialement, Jean-Daniel TISSOT

Administrateur Systèmes et Réseaux
Tel: +33 3 81 666 440 Fax: +33 3 81 666 568

Chrono-environnement

Laboratoire Chrono-environnement

16, Route de Gray
25 030 BESANÇON Cedex

Plan et Accès

___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Roland Kammerer]

On Tue, Mar 13, 2018 at 04:57:34PM +0100, Jdt wrote:
> Le 13/03/2018 à 16:21, Roland Kammerer a écrit :
> 
> > On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
> >> Hi,
> >>
> >> We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
> >> I choose to reinitialize this cluster with Proxmox V5 (stretch) with
> >> drbd from Linbit.
> >>
> >> I install Proxmox from the ISO image on the three nodes. Proxmox nodes
> >> are not associated yet in a cluster.
> >> I install drbdmanage-proxmox as indicated in
> >> https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
> >> My network interfaces are well configured for drbd.
> >> On the first node I create the drbdpool volume group as indicated in
> >> https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
> >>
> >> But it's failed in initialization :
> >> root@dmz-pve1:~ # drbdmanage init 192.168.2.10
> >>
> >> You are going to initialize a new drbdmanage cluster.
> >> CAUTION! Note that:
> >>   * Any previous drbdmanage cluster information may be removed
> >>   * Any remaining resources managed by a previous drbdmanage installation
> >>     that still exist on this system will no longer be managed by drbdmanage
> >>
> >> Confirm:
> >>
> >>   yes/no: yes
> >> Error: External command failed:
> >> drbdsetup new-resource .drbdctrl 0
> >> Command output:
> >>   (stdout)
> >> Command exited with exit_code 20
> >>
> >> Initialization failed
> > Failing at that early stage usually means that the wrong module is
> > loaded. cat /proc/drbd
> root@dmz-pve1:~ # cat /proc/drbd
> version: 8.4.7 (api:1/proto:86-101)
> srcversion: 2DCC561E7F1E3D63526E90D

Then I was right. drbdmanage always was drbd9 only. There should be a
drbd-dkms module in our proxmox repo. Install that one and make sure it
is loaded.

Regards, rck
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Jdt]

Le 13/03/2018 à 16:21, Roland Kammerer a écrit :

> On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
>> Hi,
>>
>> We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
>> I choose to reinitialize this cluster with Proxmox V5 (stretch) with
>> drbd from Linbit.
>>
>> I install Proxmox from the ISO image on the three nodes. Proxmox nodes
>> are not associated yet in a cluster.
>> I install drbdmanage-proxmox as indicated in
>> https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
>> My network interfaces are well configured for drbd.
>> On the first node I create the drbdpool volume group as indicated in
>> https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
>>
>> But it's failed in initialization :
>> root@dmz-pve1:~ # drbdmanage init 192.168.2.10
>>
>> You are going to initialize a new drbdmanage cluster.
>> CAUTION! Note that:
>>   * Any previous drbdmanage cluster information may be removed
>>   * Any remaining resources managed by a previous drbdmanage installation
>>     that still exist on this system will no longer be managed by drbdmanage
>>
>> Confirm:
>>
>>   yes/no: yes
>> Error: External command failed:
>> drbdsetup new-resource .drbdctrl 0
>> Command output:
>>   (stdout)
>> Command exited with exit_code 20
>>
>> Initialization failed
> Failing at that early stage usually means that the wrong module is
> loaded. cat /proc/drbd
root@dmz-pve1:~ # cat /proc/drbd
version: 8.4.7 (api:1/proto:86-101)
srcversion: 2DCC561E7F1E3D63526E90D

>> In /etc/drbdmanaged.cfg, I have
> You should not.
Ok, I reinstall

root@dmz-pve1:~ # modinfo drbd
filename:   /lib/modules/4.13.13-6-pve/kernel/drivers/block/drbd/drbd.ko
alias:  block-major-147-*
license:    GPL
version:    8.4.7
description:    drbd - Distributed Replicated Block Device v8.4.7
author: Philipp Reisner , Lars Ellenberg

srcversion: 2DCC561E7F1E3D63526E90D
depends:    lru_cache,libcrc32c
retpoline:  Y
intree: Y
name:   drbd
vermagic:   4.13.13-6-pve SMP mod_unload modversions
parm:   minor_count:Approximate number of drbd devices (1-255)
(uint)
parm:   disable_sendpage:bool
parm:   allow_oos:DONT USE! (bool)
parm:   proc_details:int
parm:   usermode_helper:string

root@dmz-pve1:~ # cat /etc/drbdmanaged.cfg
[LOCAL]
storage-plugin = drbdmanage.storage.lvm_thinlv.LvmThinLv
force=1
root@dmz-pve1:~ # lsmod | grep drbd
drbd  360448  0
lru_cache  16384  1 drbd
libcrc32c  16384  2 dm_persistent_data,drbd



>
> If you don't have "force" set, most of them are ignored. Having options
> per node, per config files isn't possible for years. As I don't see
> "force", but your "custom options", that indicates to me that you did
> not use our version of drbdmanage for proxmox, which sets up
> drbdmanaged.cfg correctly, or you just overwrote what was in it with
> what you thought is a good idea. Unfortunately, it was not a good
> idea...
> Regards, rck
> ___
> drbd-user mailing list
> drbd-user@lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
Regards, JdT
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user

Re: [DRBD-user] Unable to init a node

[Roland Kammerer]

On Tue, Mar 13, 2018 at 04:07:17PM +0100, Jdt wrote:
> Hi,
> 
> We had a three nodes Proxmox cluster in V4 (jessie) with drbd from Proxmox.
> I choose to reinitialize this cluster with Proxmox V5 (stretch) with
> drbd from Linbit.
> 
> I install Proxmox from the ISO image on the three nodes. Proxmox nodes
> are not associated yet in a cluster.
> I install drbdmanage-proxmox as indicated in
> https://docs.linbit.com/docs/users-guide-9.0/#ch-proxmox
> My network interfaces are well configured for drbd.
> On the first node I create the drbdpool volume group as indicated in
> https://docs.linbit.com/docs/users-guide-9.0/#s-dm-init-cluster
> 
> But it's failed in initialization :
> root@dmz-pve1:~ # drbdmanage init 192.168.2.10
> 
> You are going to initialize a new drbdmanage cluster.
> CAUTION! Note that:
>   * Any previous drbdmanage cluster information may be removed
>   * Any remaining resources managed by a previous drbdmanage installation
>     that still exist on this system will no longer be managed by drbdmanage
> 
> Confirm:
> 
>   yes/no: yes
> Error: External command failed:
> drbdsetup new-resource .drbdctrl 0
> Command output:
>   (stdout)
> Command exited with exit_code 20
> 
> Initialization failed

Failing at that early stage usually means that the wrong module is
loaded. cat /proc/drbd

> In /etc/drbdmanaged.cfg, I have

You should not.

> [GLOBAL]
> 
> # Options you can set with their default value in the GLOBAL section,
> per Site, or per Node:
> # ~~~
> 
> loglevel = INFO
> 
> [Site:DMZ]
> 
> loglevel = WARN
> 
> # Nodes available in this view:
> # ~
> #
> # You can also specify the keyword "site" in node sections
> # Example: [Node:nodeA]
> 
> [Node:dmz-pve1]
> 
> site = DMZ
> 
> storage-plugin = drbdmanage.storage.lvm_thinlv.LvmThinLv
> drbdctrl-vg = drbdpool
> 
> [Plugin:ThinLv]
> pool-name = drbdpool

If you don't have "force" set, most of them are ignored. Having options
per node, per config files isn't possible for years. As I don't see
"force", but your "custom options", that indicates to me that you did
not use our version of drbdmanage for proxmox, which sets up
drbdmanaged.cfg correctly, or you just overwrote what was in it with
what you thought is a good idea. Unfortunately, it was not a good
idea...

Regards, rck
___
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user