Koozali.org: home of the SME Server

SME Server 10 known issues

Offline Jean-Philippe Pialasse

  • *
  • 2,844
  • +11/-0
  • aka Unnilennium
    • http://smeserver.pialasse.com
SME Server 10 known issues
« on: March 24, 2021, 08:16:31 PM »
#1 I can not ssh/rsync/scp/sshfs from my SME9 to SME10 to migrate.

A the reason is we enforced only stornger cipher, and default in /etc/ssh/ssh_config. Short way is to add a supported cipher on both systems :aes256-ctr, aes192-ctr or aes128-ctr
Code: [Select]
[root@mySME9]#ssh root@mysme10 -c aes256-ctr
Code: [Select]
[root@mySME9]#sshfs -o Ciphers=aes256-ctr root@mysme10:/media/extra /mnt/backup/
Code: [Select]
[root@mySME9]#rsync  -e "ssh -c aes256-ctr" -av /var/spool/spamd/ root@mysme10:/var/spool/spamd/
Code: [Select]
[root@mySME9]#scp -c aes256-ctr  /etc/zabbix/zabbix_agentd.conf root@5.39.81.171:/etc/zabbix
#2 after I restore my data from my SME9 to my SME10 , I can not access to the web server / server-manager

Remove your templates-custom for httpd. There are chances you have a fragment with reference to php values for php module for httpd. We do not use this anymore, and it creates a fatal error on start of httpd
Code: [Select]
mv /etc/e-smith/templates-custom/etc/httpd/conf/httpd.conf /root/
 systemctl restart httpd-e-smith.service

you can generally debug issues with httpd by issuing the following command, this will output what is wrong in the file
Code: [Select]
httpd -tand you can then to a a search with the pattern of the error pointed to find what fragment seems problematic
Code: [Select]
grep -r "pattern"  /etc/e-smith/templates-custom/etc/httpd/conf/httpd.conf


#3 I might have restored template-custom added by a contrib (that should not have done that)

if you had smeserver-php-scl or smeserver-denyhosts or  smeserver-dhcpmanager you will one of those, please do

3.1- unable to access to web server or server-manager
Code: [Select]
rm -f /etc/e-smith/templates-custom/etc/httpd/conf/httpd.conf/20LoadModule80PHP
expand-template /etc/httpd/conf/httpd.conf
systemctl restart httpd-e-smith

3-2- probably no symptoms, but hsould be removed
Code: [Select]
rm -rf /etc/e-smith/templates-custom/etc/hosts.allow/sshd
expand-template /etc/hosts.allow

3.3- dhcp not working
Code: [Select]
rm -rf /etc/e-smith/templates-custom/etc/dhcpd.conf
expand-template /etc/dhcpd.conf
sysetmctl restart dhcpd.service

#4 I can not access to my ibay from my windows XP.

Windows XP and Windows 7 are software no longer supported by Microsoft, you are encouraged to migrate. That said, they are expecting the SMB1 protocol, and SME 10 uses SMB2 and SMB3. A workaround could be to allow SMB1 on the server, but this is at your own risks.


#5 I can navigate to my server using its ip, but the netbios name does not work, and it does not show in my network.

Netbios is deprecated and is part of the SMB1 protocol. You can workaround that using the contrib WSSD See https://wiki.contribs.org/Wsdd


#6 I am not able to update - I get some el6 packages proposed to be part of the updates.

There are chances you have some old el6 centos6 rhel6 repos still configured. Start with this :
Code: [Select]
yum update smeserver-yum --disablerepo=* --enablerepo=smeos,smeupdates
signal-event yum-modify
If the issue remains, then check your configured repo, remove those with reference to 6. And install them using the extrarepositories packages https://wiki.contribs.org/Extrarepositories
Code: [Select]
db yum_repositories show

#7 I am not able to update and yum seems to try to resolve to ip v6.

You have network connectivity issue or most probably dns issue. start with simple tests to more complexes. First you need to use plain ip, to exclude dns issues.

Do I have access to my ISP or to my local gateway in case something else is doing the gateway on my LAN
Code: [Select]
ping your.gate.way.ip

Do I have access to the internet. let's ping a known IP.
Code: [Select]
# ping 1.1.1.1
PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
64 bytes from 1.1.1.1: icmp_seq=1 ttl=58 time=6.11 ms

Are DNS working ?
Code: [Select]
# dig google.com +short
16.58.213.174
# dig mirror.centos.org +short
212.69.166.138
Then fix your network /dns


#8 Impossible to install got error ValueError: name already in use.

this is because you already have a SME 9 (most probably) installation on your disk with LVM.
To avoid this, either boot in rescue mode to wipe the partition, either use the partitioning  tool at screen 2 to create your own partition. Beware auto partitioning at this stage is not SME with raid or other options you had chosen but default CentOS ie EFI if needed, boot and one huge LVM with all the remaining space on ALL disks include


#9 I do not have the default raid 1 after the server installed as expected, while I have set default partitioning it the gui

Beware auto partitioning in the installer GUI is not SME with raid or other options you had chosen at boot time but default CentOS ie EFI if needed, boot and one huge LVM with all the remaining space on ALL disks include. Going in the partitioning tool will erase all previous settings chosen at boot time of installer  for SME.
« Last Edit: March 30, 2021, 05:49:43 PM by Jean-Philippe Pialasse »