Restore of shared folder failed

@davidep I had a much more successful restore. I actually installed file sharing and created a share before the backup, and the ad server was restored… this was pretty well confirmed because I had another member vm running that was joined, made a fresh backup of the ad vm, then destroyed it, started with a fresh vm with the current files in nethserver-testing installed (see screenshot, I see you made some changes with the available packages), While the fresh vm was updating I checked the member vm, it was lacking in any domain info as expected, once the restore was finished, the member vm then re-joined the ad server. Good.
But, a share I created… was not recreated, it does not exist in ibay, see error;

Jan 13 13:53:12 server7c3 smbd_audit: [2017/01/13 13:53:12.382665,  0] ../source3/smbd/service.c:808(make_connection_snum)
Jan 13 13:53:12 server7c3 smbd_audit:  canonicalize_connect_path failed for service files-test, path /var/lib/nethserver/ibay/files-test

Jan 13 13:53:04 server7c3 smbd[2363]: [2017/01/13 13:53:04.941992,  0] ../source3/smbd/service.c:808(make_connection_snum)
Jan 13 13:53:04 server7c3 smbd[2363]:  canonicalize_connect_path failed for service bob@neth.test.local, path /var/lib/nethserver/home/bob

Also… I did not have nextcloud installed in this vm, but I did have that php_scl package installed… php was set for defaut (5.4), that package… as expected, was not re-installed.
I just want to make a note of this because we need to make sure that happens because down the road if we’re using php scl to have nextcloud 11 running then obviously we’ll want that environment restored properly as well.

2 Likes

@davidep possibly relevant logged success and fails

Jan 13 13:31:31 server7c3 esmith::event[4139]: Action: /etc/e-smith/events/post-restore-config/S40nethserver-dc-post-restore SUCCESS [0.11747]
Jan 13 13:31:31 server7c3 esmith::event[4139]: Action: /etc/e-smith/events/post-restore-config/S40nethserver-sssd-restore-tdb SUCCESS [0.076089]


Jan 13 13:31:51 server7c3 esmith::event[4591]: Event: interface-update SUCCESS
Jan 13 13:31:51 server7c3 esmith::event[4139]: Action: /etc/e-smith/events/post-restore-config/S70interface-update-cond SUCCESS [19.766965]


Jan 13 13:32:46 server7c3 systemd: [/usr/lib/systemd/system/microcode.service:10] Trailing garbage, ignoring.
Jan 13 13:32:46 server7c3 systemd: microcode.service lacks both ExecStart= and ExecStop= setting. Refusing.
Jan 13 13:32:46 server7c3 systemd: Reloading.
Jan 13 13:32:46 server7c3 systemd: [/usr/lib/systemd/system/microcode.service:10] Trailing garbage, ignoring.
Jan 13 13:32:46 server7c3 systemd: microcode.service lacks both ExecStart= and ExecStop= setting. Refusing.
Jan 13 13:32:46 server7c3 systemd: Reloading.
Jan 13 13:32:46 server7c3 systemd: [/usr/lib/systemd/system/microcode.service:10] Trailing garbage, ignoring.
Jan 13 13:32:46 server7c3 systemd: microcode.service lacks both ExecStart= and ExecStop= setting. Refusing.
Jan 13 13:32:46 server7c3 esmith::event[8501]: Action: /etc/e-smith/events/runlevel-adjust/S20runlevel-adjust **FAILED**: 1 [1.319047]
Jan 13 13:32:46 server7c3 esmith::event[8501]: Event: runlevel-adjust FAILED
Jan 13 13:32:46 server7c3 esmith::event[4139]: Action: /etc/e-smith/events/post-restore-config/S70system-adjust **FAILED**: 1 [55.486312]
Jan 13 13:32:46 server7c3 esmith::event[4139]: Event: post-restore-config **FAILED**
Jan 13 13:33:01 server7c3 nmbd[7857]: [2017/01/13 13:33:01.490741,  0] ../source3/nmbd/nmbd_become_lmb.c:397(become_local_master_stage2)
Jan 13 13:33:01 server7c3 nmbd[7857]:  *****
Jan 13 13:33:01 server7c3 nmbd[7857]:  
Jan 13 13:33:01 server7c3 nmbd[7857]:  Samba name server SERVER7C3 is now a local master browser for workgroup NETH on subnet 192.168.124.210

Thank you for your valuable test, @fasttech!

Does anybody else confirm it?

I suggest splitting a separate thread on this, explaining the steps to reproduce it. /cc @stephdl

After restore-config I see the share I had on the original server under Shared folders, but it’s obviously empty (restore-data is needed).
After restore-data, everything works as expected: I can see files in the share, add files, etc.

1 Like

For an app designed in a module like nextcloud It would preferable that the developer uses the rh-php56-php of the centos scl, they are activated from start, of course if you want to install manually a web application, even one rpm based, then nethserver-php-scl makes senses.

Something to add for NS7, nethserver-php-scl is a pluginrow of nethserver-virtualhosts, nothing to do with the shared folders like in NS6…of course I can be the guilty of this bug :stuck_out_tongue:

Concerning the restoration of all installed rpm, it could be a good feature, but with the risks to get a failing restoration, it is the job of the sysadmin to recreate his environment. We can help him by saving the yum.repo and the list of rpm previously installed.

3 posts were split to a new topic: Support automatic RPM reinstall during restore procedure

@davidep @filippo_carletti I was able to recover the share and files using restore data, my bad.

2 Likes

Do you see any other ongoing issue here? Can we close this bug?

@davidep well, tbh I would prefer to do a lot more testing… because of how important disaster recovery is… but yes, I think it’s fixed and we can close it.

1 Like

Of course, new tests are welcome! I was meaning issues regarding this topic :wink:

oh, this is a split out, new topic… so for this, it’s only a bug in the tester. :blush: