[DRBD-user] DRBD 9 on pve: pending actions after container creation

Michele Rossetti rossetti at sardi.it
Thu Apr 6 10:37:56 CEST 2017

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


After some working days in production (3 PVE server with DRBD9), I 
found some KVM inconsistent on 2 secondary nodes (I don't use 
containers).

After upgrading yesterday to last PVE and DRBD:
root at mpve2:~# uname -a
Linux mpve2 4.4.49-1-pve #1 SMP PVE 4.4.49-86 (Thu, 30 Mar 2017 
08:39:20 +0200) x86_64 GNU/Linux

and rebooting the nodes, the system has done sync as here below:

root at mpve2:~# drbd-overview
   0:.drbdctrl/0      Connected(3*) Prim(mpve2)/Seco(mpve3,mpve1) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
   1:.drbdctrl/1      Connected(3*) Prim(mpve2)/Seco(mpve1,mpve3) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
102:vm-102-disk-1/0  Connected(3*) Secondary(3*) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
103:vm-103-disk-1/0  Connected(3*) Secondary(3*) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
104:vm-104-disk-1/0  Connected(3*) Prim(mpve2)/Seco(mpve1,mpve3) 
UpTo(mpve2)/UpTo(mpve3,mpve1)
116:vm-100-disk-1/0  Connected(3*) Prim(mpve2)/Seco(mpve1,mpve3) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
117:vm-101-disk-1/0  Connected(3*) Prim(mpve2)/Seco(mpve3,mpve1) 
UpTo(mpve2)/UpTo(mpve1,mpve3)
121:vm-106-disk-1/0  Connected(3*) Prim(mpve2)/Seco(mpve3,mpve1) 
UpTo(mpve2)/UpTo(mpve3,mpve1)
122:vm-107-disk-1/0  Connected(3*) Prim(mpve2)/Seco(mpve1,mpve3) 
UpTo(mpve2)/UpTo(mpve3,mpve1)
123:vm-108-disk-1/0  Connected(3*) Secondary(3*) 
UpTo(mpve2)/UpTo(mpve3,mpve1)

root at mpve2:~# drbdsetup status
.drbdctrl role:Primary
   volume:0 disk:UpToDate
   volume:1 disk:UpToDate
   mpve1 role:Secondary
     volume:0 peer-disk:UpToDate
     volume:1 peer-disk:UpToDate
   mpve3 role:Secondary
     volume:0 peer-disk:UpToDate
     volume:1 peer-disk:UpToDate

vm-100-disk-1 role:Primary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-101-disk-1 role:Primary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-102-disk-1 role:Secondary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-103-disk-1 role:Secondary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-104-disk-1 role:Primary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-106-disk-1 role:Primary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-107-disk-1 role:Primary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate

vm-108-disk-1 role:Secondary
   disk:UpToDate
   mpve1 role:Secondary
     peer-disk:UpToDate
   mpve3 role:Secondary
     peer-disk:UpToDate


But unfortunatly, checking with drbdmanage a there are some resources 
with pending actions:

root at mpve2:~# drbdmanage a
+----------------------------------------------------------------------------------------------------------+
| Node  | Resource      | Vol ID | 
|                               State |
|----------------------------------------------------------------------------------------------------------|
| mpve1 | vm-108-disk-1 |      * | 
|         pending actions: commission |
| mpve1 | vm-108-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve1 | vm-107-disk-1 |      * | 
|         pending actions: commission |
| mpve1 | vm-107-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve1 | vm-106-disk-1 |      * | 
|         pending actions: commission |
| mpve1 | vm-106-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve1 | vm-104-disk-1 |      * | 
|                                  ok |
| mpve1 | vm-103-disk-1 |      * | 
|                                  ok |
| mpve1 | vm-102-disk-1 |      * | 
|                                  ok |
| mpve1 | vm-101-disk-1 |      * | 
|                                  ok |
| mpve1 | vm-100-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-108-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-107-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-106-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-104-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-103-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-102-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-101-disk-1 |      * | 
|                                  ok |
| mpve2 | vm-100-disk-1 |      * | 
|                                  ok |
| mpve3 | vm-108-disk-1 |      * | 
|         pending actions: commission |
| mpve3 | vm-108-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve3 | vm-107-disk-1 |      * | 
|         pending actions: commission |
| mpve3 | vm-107-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve3 | vm-106-disk-1 |      * | 
|         pending actions: commission |
| mpve3 | vm-106-disk-1 |      0 | 
| pending actions: commission, attach |
| mpve3 | vm-104-disk-1 |      * | 
|                                  ok |
| mpve3 | vm-103-disk-1 |      * | 
|                                  ok |
| mpve3 | vm-102-disk-1 |      * | 
|                                  ok |
| mpve3 | vm-101-disk-1 |      * | 
|                                  ok |
| mpve3 | vm-100-disk-1 |      * | 
|                                  ok |
+----------------------------------------------------------------------------------------------------------+

Can someone explain to me why there is this difference between 
<drbdsetup status> and <drbdmanage a> and how to clean the system?

I had to understand too why the system, that was clean at startup of 
VM's, after a few days was gone by itself in inconsistent state on 3 
VM's.

Maybe DRBD9 is really not yet ready for production use... or I'm gone wrong?

Thanks,
Michele


>Il 29/03/2017 16:00, Roberto Resoli ha scritto:
>>  Il 26/03/2017 11:19, Roberto Resoli ha scritto:
>>>  I tried to create a new resource bypassing pve storage plugin, using
>>>  drbmanage:
>>>
>>>  drbdmanage add-volume prova 200MB --deploy 3
>>>
>>>  I get a clean "drbdmanage a" status on the same node, and "pending
>>>  actions" on both peer nodes.
>>
>>  I tried to restore a vm (drbd resouce vm-105-disk-1), same results.
>>
>>  This time I can provide the relevant syslog portions at DEBUG level,
>>  both for pve2 node (the target of restore) and pve3 (the node which
>>  exibits pending actions). I attach them to this message.
>
>Update: After having updated all three nodes with very latest kernel
>
>uname -a
>Linux pve1 4.4.49-1-pve #1 SMP PVE 4.4.49-86 (Thu, 30 Mar 2017 08:39:20
>+0200) x86_64 GNU/Linux
>
>all "pending actions" in "drbdmanage a" output disappeared.
>
>The assignment status is now clean.
>
>bye,
>rob
>_______________________________________________
>drbd-user mailing list
>drbd-user at lists.linbit.com
>http://lists.linbit.com/mailman/listinfo/drbd-user


-- 
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
MICRO srl
Informatica e Telecomunicazioni - Internet services - Web sites
      
       	Michele Rossetti

sede legale: via Raffa Garzia 7   09126 Cagliari (Italy)
sede operativa: viale Marconi 222  09131 Cagliari
Ph. +39 070 400240  Fax +39 070 4526207

MKM-REG
Web:  http://www.microsrl.com     http://www.sardi.it
E-mail: microsrl at microsrl.com
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""



More information about the drbd-user mailing list