[DRBD-user] linstor issues
Adam Goryachev
mailinglists at websitemanagers.com.au
Fri Sep 18 08:57:37 CEST 2020
Hi all,
I've finally had another crack at this, and eventually got it working.
I think I had all of the following issues at various stages, so just
mentioning them here for other future people who find they have similar
issues.
1) Initially, I had a firewall blocking connections between servers. I
fixed this early on, but not sure if it messed up some of the config and
never recovered.
2) Somehow, some of the servers ended up with old 8.4 kernel module
version installed. I think this was related to installing new kernels
from debian, but for some reason it wouldn't update matching header
files, and so dkms couldn't recompile the module for the new kernel.
Thus, reverting to the in-kernel 8.4 version.
3) Finally, I think the hardest issue to track down was that there was a
copy of drbdadm and associated drbd-utils installed under /usr/local and
these were taking priority over the package installed version. Thus,
drbdadm couldn't find the drbd resource config files because linstor was
putting them in /etc/drbd.d and drbdadm was looking in /usr/local/etc/drbd.d
Question: Is this something that linstor should be aware of (ie, storing
the resource files in the same location as the installed/used drbdadm
executable expects them to go)? I guess it's a pretty rare issue, but
curious if it could be avoided easily.
After a few complete wipe/reinstalls, I found and fixed these issues,
and now looks like I have a working installation.
linstor node list
╭──────────────────────────────────────────────────────────╮
┊ Node ┊ NodeType ┊ Addresses ┊ State ┊
╞══════════════════════════════════════════════════════════╡
┊ castle ┊ SATELLITE ┊ 192.168.5.204:3366 (PLAIN) ┊ Online ┊
┊ san5 ┊ SATELLITE ┊ 192.168.5.205:3366 (PLAIN) ┊ Online ┊
┊ san6 ┊ SATELLITE ┊ 192.168.5.206:3366 (PLAIN) ┊ Online ┊
╰──────────────────────────────────────────────────────────╯
linstor storage-pool list
╭───────────────────────────────────────────────────────────────────────────────────────────────────────────╮
┊ StoragePool ┊ Node ┊ Driver ┊ PoolName ┊ FreeCapacity ┊
TotalCapacity ┊ CanSnapshots ┊ State ┊
╞═══════════════════════════════════════════════════════════════════════════════════════════════════════════╡
┊ DfltDisklessStorPool ┊ castle ┊ DISKLESS ┊ ┊
┊ ┊ False ┊ Ok ┊
┊ DfltDisklessStorPool ┊ san5 ┊ DISKLESS ┊ ┊
┊ ┊ False ┊ Ok ┊
┊ DfltDisklessStorPool ┊ san6 ┊ DISKLESS ┊ ┊
┊ ┊ False ┊ Ok ┊
┊ pool_hdd ┊ castle ┊ LVM ┊ vg_hdd ┊ 2.95 TiB
┊ 3.44 TiB ┊ False ┊ Ok ┊
┊ pool_hdd ┊ san5 ┊ LVM ┊ vg_hdd ┊ 3.87 TiB
┊ 4.36 TiB ┊ False ┊ Ok ┊
┊ pool_hdd ┊ san6 ┊ LVM ┊ vg_hdd ┊ 1.26 TiB
┊ 1.75 TiB ┊ False ┊ Ok ┊
╰───────────────────────────────────────────────────────────────────────────────────────────────────────────╯
Thank you for those that replied and helped out along the way, I'm
finally feeling a lot more confident in moving forward with this
project. I'll no doubt be asking a few more questions along the way, but
hopefully there won't be anything quite as unexpected as this (#3)
happening.
Regards,
Adam
On 14/7/20 16:43, Adam Goryachev wrote:
>
> I'm having another crack at this, I think it will be worth it once it
> works.
>
> Firstly, another documentation error:
>
> https://www.linbit.com/drbd-user-guide/linstor-guide-1_0-en/#s-using_the_linstor_client
>
>> In case anything goes wrong with the storage pool’s VG/zPool, e.g.
>> the VG having been renamed or somehow became invalid you can delete
>> the storage pool in LINSTOR with the following command, given that
>> only resources with all their volumes in the so-called ‘lost’ storage
>> pool are attached. This feature is available since LINSTOR v0.9.13.
>>
>> # linstor storage-pool lost alpha pool_ssd
> linstor storage-pool lost castle vg_hdd
> usage: linstor storage-pool [-h]
> {create, delete, list, list-properties,
> set-property} ...
> linstor storage-pool: error: argument {create, delete, list,
> list-properties, set-property}: invalid choice: 'lost' (choose from
> 'create', 'c', 'delete', 'd', 'list', 'l', 'list-properties', 'lp',
> 'set-property', 'sp')
>
> Changing to use delete instead of lost:
>
> castle:~# linstor storage-pool delete castle vg_hdd
> ERROR:
> Description:
> Storage pool definition 'vg_hdd' not found.
> Cause:
> The specified storage pool definition 'vg_hdd' could not be found
> in the database
> Correction:
> Create a storage pool definition 'vg_hdd' first.
> Details:
> Node: castle, Storage pool name: vg_hdd
> Show reports:
> linstor error-reports show 5F0D500C-00000-000000
> castle:~# linstor storage-pool list
> ╭───────────────────────────────────────────────────────────────────────────────────────────────────────────╮
> ┊ StoragePool ┊ Node ┊ Driver ┊ PoolName ┊ FreeCapacity ┊
> TotalCapacity ┊ CanSnapshots ┊ State ┊
> ╞═══════════════════════════════════════════════════════════════════════════════════════════════════════════╡
> ┊ DfltDisklessStorPool ┊ castle ┊ DISKLESS ┊ ┊
> ┊ ┊ False ┊ Ok ┊
> ┊ DfltDisklessStorPool ┊ san5 ┊ DISKLESS ┊ ┊
> ┊ ┊ False ┊ Ok ┊
> ┊ DfltDisklessStorPool ┊ san6 ┊ DISKLESS ┊ ┊
> ┊ ┊ False ┊ Ok ┊
> ┊ pool ┊ castle ┊ LVM ┊ vg_hdd ┊ 2.95 TiB
> ┊ 3.44 TiB ┊ False ┊ Ok ┊
> ┊ pool ┊ san5 ┊ LVM ┊ vg_hdd ┊ 3.87 TiB
> ┊ 4.36 TiB ┊ False ┊ Ok ┊
> ┊ pool ┊ san6 ┊ LVM ┊ vg_ssd ┊ 1.26 TiB
> ┊ 1.75 TiB ┊ False ┊ Ok ┊
> ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────╯
>
> I was hoping I could just remove the storage pool from castle (since
> it doesn't seem to be working properly), and then destroy it,
> re-create it, and then re-add it and see if that solves the problem.
> However, while it seems to exist, it also doesn't (can't delete it).
>
> Possibly part of the cause of my original problem is that I have a
> script that automatically creates a snapshot for each LV, and this
> created a snapshot of testvm1_00000 named
> backup_testvm1_00000_blahblah.... I've now manually deleted that, and
> fixed my script to avoid messing with the VG allocated to linstor, but
> so far, there is no change in the current status (as per below).
>
> Would appreciate any suggestions on what might be going wrong, and/or
> how to fix it?
>
> Regards,
> Adam
>
>
>
> On 24/6/20 11:46, Adam Goryachev wrote:
>>
>>
>> On 23/6/20 21:53, Gábor Hernádi wrote:
>>> Hi,
>>>
>>> apparently something is quite broken... maybe it's somehow your
>>> setup or environment, I am not sure...
>>>
>>> linstor resource list
>>> ╭────────────────────────────────────────────────────────────────────────────╮
>>> ┊ ResourceName ┊ Node ┊ Port ┊ Usage ┊
>>> Conns ┊ State ┊
>>> ╞════════════════════════════════════════════════════════════════════════════╡
>>> ┊ testvm1 ┊ castle ┊ 7000 ┊ ┊ ┊
>>> Unknown ┊
>>> ┊ testvm1 ┊ san5 ┊ 7000 ┊ ┊ ┊
>>> Unknown ┊
>>> ┊ testvm1 ┊ san6 ┊ 7000 ┊ Unused ┊
>>> Connecting(san5,castle) ┊ UpToDate ┊
>>> ╰────────────────────────────────────────────────────────────────────────────╯
>>>
>>> This looks like some kind of network issues.
>>>
>>> # linstor storage-pool list --groupby Size
>>>
>>> However, the second command produces a usage error
>>> (documentation bug perhaps).
>>>
>>>
>>> Thanks for reporting, we will look into this.
>>>
>>> WARNING:
>>> Description:
>>> No active connection to satellite 'san5'
>>> Details:
>>> The controller is trying to (re-) establish a connection to
>>> the satellite. The controller stored the changes and as soon the
>>> satellite is connected, it will receive this update.
>>>
>>>
>>> So Linstor has obviously no connection to satellite 'san5'.
>>>
>>> [95078.599813] drbd testvm1 castle: conn( Unconnected ->
>>> Connecting )
>>> [95078.604454] drbd testvm1 san5: conn( Unconnected -> Connecting )
>>>
>>>
>>> ... and DRBD apparently also has troubles connecting...
>>>
>>> linstor n l
>>> ╭───────────────────────────────────────────────────────────╮
>>> ┊ Node ┊ NodeType ┊ Addresses ┊ State ┊
>>> ╞═══════════════════════════════════════════════════════════╡
>>> ┊ castle ┊ SATELLITE ┊ 192.168.5.204:3366
>>> <http://192.168.5.204:3366> (PLAIN) ┊ Unknown ┊
>>> ┊ san5 ┊ SATELLITE ┊ 192.168.5.205:3366
>>> <http://192.168.5.205:3366> (PLAIN) ┊ Unknown ┊
>>> ┊ san6 ┊ SATELLITE ┊ 192.168.5.206:3366
>>> <http://192.168.5.206:3366> (PLAIN) ┊ Unknown ┊
>>> ╰───────────────────────────────────────────────────────────╯
>>>
>>>
>>> Now this is really strange. I will spare you with some details, but
>>> I assume you have triggered some bad exception in Linstor which
>>> somehow killed a necessary thread.
>>> You should check
>>> linstor err list
>>> and see if you can find some related error reports.
>>> Also, restarting the controller might help you here.
>>>
>> Thank you!
>>
>> linstor err list showed a list of errors, but the contents didn't
>> make a lot of sense to me. Let me know if you are interested in them,
>> and I can send them.
>>
>> I did a systemctl restart linstor-controller.service on san6, and
>> things started looking much better.
>>
>> linstor n l
>> ╭──────────────────────────────────────────────────────────╮
>> ┊ Node ┊ NodeType ┊ Addresses ┊ State ┊
>> ╞══════════════════════════════════════════════════════════╡
>> ┊ castle ┊ SATELLITE ┊ 192.168.5.204:3366 (PLAIN) ┊ Online ┊
>> ┊ san5 ┊ SATELLITE ┊ 192.168.5.205:3366 (PLAIN) ┊ Online ┊
>> ┊ san6 ┊ SATELLITE ┊ 192.168.5.206:3366 (PLAIN) ┊ Online ┊
>> ╰──────────────────────────────────────────────────────────╯
>>
>> So, all nodes agree that they are now online and talking to each
>> other. I assume this proves there is no network issues.
>>
>> linstor resource list
>> ╭─────────────────────────────────────────────────────────────────────────────────╮
>> ┊ ResourceName ┊ Node ┊ Port ┊ Usage ┊ Conns ┊ State ┊
>> ╞═════════════════════════════════════════════════════════════════════════════════╡
>> ┊ testvm1 ┊ castle ┊ 7000 ┊ ┊ ┊ Unknown ┊
>> ┊ testvm1 ┊ san5 ┊ 7000 ┊ Unused ┊ Connecting(castle) ┊
>> SyncTarget(12.67%) ┊
>> ┊ testvm1 ┊ san6 ┊ 7000 ┊ Unused ┊ Connecting(castle)
>> ┊ UpToDate ┊
>> ╰─────────────────────────────────────────────────────────────────────────────────╯
>>
>> From this, it looks like san6 (the controller) thinks it has the up
>> to date data, probably based on the fact it was created there first
>> or something. The data is syncing to san5 (in progress, and
>> progressing steadily), so that is good also. However, castle doesn't
>> seem to be syncing/connecting.
>>
>> On castle, I see this:
>>
>> Jun 24 11:01:55 castle Satellite[7499]: 11:01:55.177 [DeviceManager]
>> ERROR LINSTOR/Satellite - SYSTEM - Failed to create meta-data for
>> DRBD volume testvm1/0 [Report number 5EF2A316-31431-000002]
>>
>> linstor err show give this:
>>
>> ERROR REPORT 5EF2A316-31431-000002
>>
>> ============================================================
>>
>> Application: LINBIT® LINSTOR
>> Module: Satellite
>> Version: 1.7.1
>> Build ID: 6760637d6fae7a5862103ced4ea0ab0a758861f9
>> Build time: 2020-05-14T13:14:11+00:00
>> Error time: 2020-06-24 11:01:55
>> Node: castle
>>
>> ============================================================
>>
>> Reported error:
>> ===============
>>
>> Description:
>> Failed to create meta-data for DRBD volume testvm1/0
>>
>> Category: LinStorException
>> Class name: VolumeException
>> Class canonical name:
>> com.linbit.linstor.storage.layer.exceptions.VolumeException
>> Generated at: Method 'createMetaData', Source
>> file 'DrbdLayer.java', Line #995
>>
>> Error message: Failed to create meta-data for
>> DRBD volume testvm1/0
>>
>> Error context:
>> An error occurred while processing resource 'Node: 'castle', Rsc:
>> 'testvm1''
>>
>> Call backtrace:
>>
>> Method Native Class:Line number
>> createMetaData N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:995
>> adjustDrbd N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:575
>> process N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:373
>> process N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:731
>> processResourcesAndSnapshots N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:300
>> dispatchResources N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:138
>> dispatchResources N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:258
>> phaseDispatchDeviceHandlers N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:896
>> devMgrLoop N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:618
>> run N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:535
>> run N java.lang.Thread:834
>>
>> Caused by:
>> ==========
>>
>> Description:
>> Execution of the external command 'drbdadm' failed.
>> Cause:
>> The external command exited with error code 1.
>> Correction:
>> - Check whether the external program is operating properly.
>> - Check whether the command line is correct.
>> Contact a system administrator or a developer if the command
>> line is no longer valid
>> for the installed version of the external program.
>> Additional information:
>> The full command line executed was:
>> drbdadm -vvv --max-peers 7 -- --force create-md testvm1/0
>>
>> The external command sent the following output data:
>>
>>
>> The external command sent the following error information:
>> no resources defined!
>>
>>
>> Category: LinStorException
>> Class name: ExtCmdFailedException
>> Class canonical name: com.linbit.extproc.ExtCmdFailedException
>> Generated at: Method 'execute', Source file
>> 'DrbdAdm.java', Line #550
>>
>> Error message: The external command 'drbdadm'
>> exited with error code 1
>>
>>
>> Call backtrace:
>>
>> Method Native Class:Line number
>> execute N
>> com.linbit.linstor.storage.layer.adapter.drbd.utils.DrbdAdm:550
>> simpleAdmCommand N
>> com.linbit.linstor.storage.layer.adapter.drbd.utils.DrbdAdm:495
>> createMd N
>> com.linbit.linstor.storage.layer.adapter.drbd.utils.DrbdAdm:262
>> createMetaData N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:923
>> adjustDrbd N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:575
>> process N
>> com.linbit.linstor.storage.layer.adapter.drbd.DrbdLayer:373
>> process N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:731
>> processResourcesAndSnapshots N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:300
>> dispatchResources N
>> com.linbit.linstor.core.devmgr.DeviceHandlerImpl:138
>> dispatchResources N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:258
>> phaseDispatchDeviceHandlers N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:896
>> devMgrLoop N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:618
>> run N
>> com.linbit.linstor.core.devmgr.DeviceManagerImpl:535
>> run N java.lang.Thread:834
>>
>>
>> END OF ERROR REPORT.
>>
>> Indeed, re-running the same command from the CLI provides the shown
>> error message:
>>
>> drbdadm -vvv --max-peers 7 -- --force create-md testvm1/0
>> no resources defined!
>>
>> Some other random status information which may or may not be relevant...
>>
>> linstor storage-pool list
>> ╭───────────────────────────────────────────────────────────────────────────────────────────────────────────╮
>> ┊ StoragePool ┊ Node ┊ Driver ┊ PoolName ┊ FreeCapacity
>> ┊ TotalCapacity ┊ CanSnapshots ┊ State ┊
>> ╞═══════════════════════════════════════════════════════════════════════════════════════════════════════════╡
>> ┊ DfltDisklessStorPool ┊ castle ┊ DISKLESS ┊ ┊
>> ┊ ┊ False ┊ Ok ┊
>> ┊ DfltDisklessStorPool ┊ san5 ┊ DISKLESS ┊ ┊
>> ┊ ┊ False ┊ Ok ┊
>> ┊ DfltDisklessStorPool ┊ san6 ┊ DISKLESS ┊ ┊
>> ┊ ┊ False ┊ Ok ┊
>> ┊ pool ┊ castle ┊ LVM ┊ vg_hdd ┊ 2.95 TiB
>> ┊ 3.44 TiB ┊ False ┊ Ok ┊
>> ┊ pool ┊ san5 ┊ LVM ┊ vg_hdd ┊ 3.87 TiB
>> ┊ 4.36 TiB ┊ False ┊ Ok ┊
>> ┊ pool ┊ san6 ┊ LVM ┊ vg_ssd ┊ 1.26 TiB
>> ┊ 1.75 TiB ┊ False ┊ Ok ┊
>> ╰───────────────────────────────────────────────────────────────────────────────────────────────────────────╯
>>
>> I've tried to restart linstor-satellite service on castle, but it
>> didn't make any difference.
>>
>> After a reboot of castle, and now I get this:
>>
>> linstor resource list
>> ╭────────────────────────────────────────────────────────────────────╮
>> ┊ ResourceName ┊ Node ┊ Port ┊ Usage ┊ Conns ┊ State ┊
>> ╞════════════════════════════════════════════════════════════════════╡
>> ┊ testvm1 ┊ castle ┊ 7000 ┊ Unused ┊ Ok ┊ Diskless ┊
>> ┊ testvm1 ┊ san5 ┊ 7000 ┊ Unused ┊ Ok ┊ SyncTarget(55.99%) ┊
>> ┊ testvm1 ┊ san6 ┊ 7000 ┊ Unused ┊ Ok ┊ UpToDate ┊
>> ╰────────────────────────────────────────────────────────────────────╯
>>
>> However, looking at the err reports, and I see the exactl same error
>> about creating the metadata on castle.
>>
>> One interesting thing is that the LV seems to have been created:
>>
>> lvs
>> /dev/drbd0: open failed: Wrong medium type
>> /dev/drbd1: open failed: Wrong medium type
>> LV VG Attr LSize Pool Origin
>> Data% Meta% Move Log Cpy%Sync Convert
>> backup_system_20200624_062513 storage swi-a-s--- 4.00g system
>> 3.06
>> system storage owi-aos--- 5.00g
>> testvm1_00000 vg_hdd -wi-a----- <500.11g
>>
>> Any suggestions on where to look next? Or what I might have done
>> wrong now?
>>
>> Regards,
>> Adam
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> Star us on GITHUB:https://github.com/LINBIT
>> drbd-user mailing list
>> drbd-user at lists.linbit.com
>> https://lists.linbit.com/mailman/listinfo/drbd-user
>
> _______________________________________________
> Star us on GITHUB: https://github.com/LINBIT
> drbd-user mailing list
> drbd-user at lists.linbit.com
> https://lists.linbit.com/mailman/listinfo/drbd-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20200918/d59ffc97/attachment-0001.htm>
More information about the drbd-user
mailing list