Questions about migration from NS7 to NS8

Hello everyone, buon anno nuovo, happy new year !
i tried today the first time to migrate my NS7 to NS8 with the migration tool and did not succeed, first thing what i recognized after i got the error at the ns7 was the circumstances it creates new apps on the ns8. The ns8 is with the whole Updates from what was possible. Is there a reason the it is not allowed to get the latest Updates without doing the migration from ns7? I have a working ns8 on Rocky 9.5 things are running like mail sogo different other apps. After the update i got new instances of apps in mail, sogo, etc. mail is at the working ns8 mail1, with migration i got mail3, same at the sogo instances. both are vm at a proxmox system, but i dont thing there is the prob. my plan is now to install the ns8 completely new and i would like to got an advice which is the better os for ns8? Rocky or Debian? How big do i need the Harddisk for a small business with around 10 Workers, Email, Kalender and a few Dokument in the Directory of the samba share? And are there things to plan before installing NS8 that the migration works? I know, a lot questions and certainly some things are also due to personal taste, but everyone has certainly had their own experiences and I would like to know whether there are any trends in the various areas, thank you very much.

Updates could be an issue during migration, see also Nextcloud migration fails if automatic updates are enabled · Issue #7070 · NethServer/dev · GitHub

You should migrate to a fresh installed NS8, see also NethServer 7 migration — NS8 documentation

Subscription is only supported with Rocky.
Rocky is an RHEL clone and has the more recent podman version 5.2.2 whereas Debian still uses 4.3.1 but that will change with next Debian release, see also Debian -- Package Search Results -- podman
Debian has the larger community and more documentation/tutorials.
I’d recommend the OS you’re more familiar with. I use both without issues.

In any case I’d recommend fast SSD/NVME disks.

If it’s about physical disk I think 2 TB are should cover your case including some free space for future but It really depends on how much (mail) data is used.
As regards virtual disks, the size can be increased later.

As it already worked there should be no issues.

1 Like

i dont know how often i said thank you in the last years, but again THANK YOU @mrmarkuz Markus for your help, time and good will.

1 Like

another short question, i installed the ns8 now fresh on a machine, is it necessary to install the apps like mail sogo etc. before migrating the ns7 or should i leave the ns8 clean without any apps?

@hucky

Leave it clean, NS8 will install the apps as required

This is “controlled” by the migration app, running on NS7.
:slight_smile:

My 2 cents
Andy

2 Likes


i did install fresh clean with out installing apps or anything else and got this error. Error migrating data

do i have to install the active directory/samba first?

echo ‘{“app”:“nethserver-mail”,“action”:“sync”}’ | /usr/bin/setsid /usr/bin/sudo /usr/libexec/nethserver/api/nethserver-ns8-migration/migration/update | jq

Log Entry:
Joined to cluster leader ns8.xxx.de
----------- start nethserver-mail Fri, 24 Jan 2025 13:22:29 +0100
Assertion failed
File “/var/lib/nethserver/cluster/actions/import-module/50import”, line 50, in
agent.assert_exp(add_module_result[‘exit_code’] == 0) # add-module is successful
[INFO] Created remote module instance imapsync2
[INFO] App nethserver-mail-getmail is bound to rsync://imapsync2@10.5.4.1:20003, waiting for task module/imapsync2/task/07b32b22-2ef7-4e3d-9df3-36eb0555ee9d
[INFO] Created remote module instance mail2
[INFO] App nethserver-mail is bound to rsync://mail2@10.5.4.1:20004, waiting for task module/mail2/task/d9be303d-f466-4889-bb20-628328c27bbc
=========== Leave cluster Fri, 24 Jan 2025 13:23:46 +0100
=========== Join cluster Fri, 24 Jan 2025 14:38:55 +0100
ns8-join: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:877)>
=========== Join cluster Fri, 24 Jan 2025 14:38:59 +0100
Joined to cluster leader ns8.xxx.de
----------- start nethserver-mail Fri, 24 Jan 2025 14:39:24 +0100
Assertion failed
File “/var/lib/nethserver/cluster/actions/import-module/50import”, line 50, in
agent.assert_exp(add_module_result[‘exit_code’] == 0) # add-module is successful
[INFO] Created remote module instance imapsync1
[INFO] App nethserver-mail-getmail is bound to rsync://imapsync1@10.5.4.1:20002, waiting for task module/imapsync1/task/236939a3-5a99-42ee-8b7a-46e8d86e514d
[INFO] Created remote module instance mail1
[INFO] App nethserver-mail is bound to rsync://mail1@10.5.4.1:20003, waiting for task module/mail1/task/741445cc-f345-491e-8374-b78b21da7637
----------- sync nethserver-mail Fri, 24 Jan 2025 15:02:14 +0100
----------- sync nethserver-mail Fri, 24 Jan 2025 15:08:50 +0100

There is an issue in Core 3.4.3 (extract-image command) that leads to add-module failures. Involved scenarios: new cluster, join cluster, migration, restore
 If two add-modules run concurrently one of them fails for a lock file issue. Reference fix: extract-image regression by DavidePrincipi · Pull Request #791 · NethServer/ns8-core · GitHub.

Possible workarounds:

  • for new cluster / join cluster errors with Core 3.4.3: reinstall NS8
  • for other cases, where installation/join ran with core <3.4.3, update to Core 3.4.4, then retry. For example, with migration, abort the failing module and start again.
2 Likes

In this case reinstall ns8 starting with core 3.4.4, or you might find more troubles ahead!

2 Likes

i did the installation today, first Rocky 9.5 and after that i installed the NS8 with curl https://raw.githubusercontent.com/NethServer/ns8-core/ns8-stable/core/install.sh | bash Link. is in that Installation Routine the Core 3.4.4 not included?
The Update from the Software Center is not possible because of:
Updates are suspended
There is an ongoing NS7 migration. Updates will resume once the migration is complete.
I would do it completeley new if then the right core is installed


i uninstall the NS8 and install it again, from what i saw at the installation it was the 3.4.4 but same error the mail and imapsynch is installed but then i got the error, if i now try to abort the migration i got this error There was an error during the abort procedure.


i am sorry but it is the same as before, which kind of a protocol could be helpful? What i did is install Rocky 9.5 from scratch to a new VM on Proxmox. After that i installed the NS8 from the Installation Guide NS8. After that i did an reboot and start the migration on NS7. It connect to the NS8 Cluster and i click Migration, after that i start the the migration to ns8 and got the Error migrating data message. if i look at the NS8 i got installed Mail and imapsync but not more.

Please attach the ns8-migration.log. You can find it from NS7 Cockpit UI, under the Migration Tool page.

File “/var/lib/nethserver/cluster/actions/import-module/50import”, line 50, in
agent.assert_exp(add_module_result[‘exit_code’] == 0) # add-module is successful
[INFO] Created remote module instance imapsync1
[INFO] App nethserver-mail-getmail is bound to rsync://imapsync1@10.5.4.1:20001, waiting for task module/imapsync1/task/96181f91-9ba2-4d1c-9dc7-21c6ed86a284
[INFO] Created remote module instance mail1
[INFO] App nethserver-mail is bound to rsync://mail1@10.5.4.1:20002, waiting for task module/mail1/task/b52a2b7e-0e5c-4735-b7d2-1b6a72dba088
----------- abort nethserver-mail Fri, 24 Jan 2025 19:38:03 +0100
=========== Leave cluster Fri, 24 Jan 2025 19:50:14 +0100
=========== Leave cluster Fri, 24 Jan 2025 19:50:21 +0100
=========== Join cluster Mon, 27 Jan 2025 15:32:08 +0100
ns8-join: <urlopen error [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:877)>
=========== Join cluster Mon, 27 Jan 2025 15:32:11 +0100
Joined to cluster leader ns8.xxx.de
----------- start nethserver-mail Mon, 27 Jan 2025 15:34:16 +0100
Assertion failed
File “/var/lib/nethserver/cluster/actions/import-module/50import”, line 50, in
agent.assert_exp(add_module_result[‘exit_code’] == 0) # add-module is successful
[INFO] Created remote module instance imapsync1
[INFO] App nethserver-mail-getmail is bound to rsync://imapsync1@10.5.4.1:20001, waiting for task module/imapsync1/task/03c20e95-74ab-4fc2-9c0e-f164f47447c6
[INFO] Created remote module instance mail1
[INFO] App nethserver-mail is bound to rsync://mail1@10.5.4.1:20002, waiting for task module/mail1/task/e92886e5-698f-434e-ac6f-f37c6411bbc0

Is the error from SOGo? As the docs states, you’ve to enable Nethforge in NS8 Settings > Repositories before starting the migration.

2 Likes

okay, will try it directly

Good morning everyone.
I have now completed the first test migration and still encountered a few problems.
When the migration was completed and everything was transferred, including the domain controller, I was still unable to join the domain. An error was always generated that the domain and the controller could not be found. The fact is that the new domain server under NS8 has the same domain names as the old server NS7 (where joining works flawlessly). Is this the problem? How can it be prevented? The migration went smoothly, all users were transferred etc. Furthermore, when installing the DNSmasc module I am told that the DNS is blocked by the Samba, is this intentional?Is there a possibility to check which ip adress, like in ns7 the new one ns8 have?

Joining on NS7 still works?

Did you go through the post migration steps for samba in NethServer 7 migration — NS8 documentation?

Yes, you can use dnsmasq for DHCP but not for DNS when port 53 is used by samba, see also NethServer project milestone 8.3

In NS7 the samba DC had another IP, in NS8 usually the host IP is used to be reachable from clients.

To get the IP, you could use

ip a

hello markus, and thank you again for your detailed help, without you I would be lost. I have carried out the migration to the end and virtually worked through the documentation. But I didn’t try to rejoin the old domain, only to join the new one after the migration. I think it may be because I may not have given the client to be connected the new server as 1 DNS? I will check this tomorrow morning, how do I get the dns entry for the new server for virtual addresses in the network? Can’t I set the entries in DNSmasq so that it is distributed in the network? the Ip address of the new server is the correct one, maybe it was indeed a DNS problem, I will find out tomorrow, I hope. But I have to reconnect the client, right? Is it not enough to boot the computer after the migration to move it to the new network? Have a nice evening and thanks again.

1 Like