Topic: Console Upgrade from 9.2 to 10- where did I go wrong.

[k_graham]

your system is expected to be slow on first hours as it is generating a dh parameter for dovecot.  could take more than one day on really slow system.

also please not that bu choosing to format with ext4 you deviated from standard install which will use xfs, and had to choose (wisely, we all hope) your partitionning, as even entering in the disk interface will result is overriding the kickstart default we provide and will replace it with either your settings or centos defaults.

your system is expected to be slow on first hours as it is generating a dh parameter for dovecot.  could take more than one day on really slow system.

also please not that bu choosing to format with ext4 you deviated from standard install which will use xfs, and had to choose (wisely, we all hope) your partitionning, as even entering in the disk interface will result is overriding the kickstart default we provide and will replace it with either your settings or centos defaults.

I deleted the previous install to the disks by overwriting with ext4. From what you say perhaps I could have left the old one in place to be overwritten or simply deleted to start with a blank non formatted disk?

 I did not change the Custom setting of the screen during install or enter it , I just checked the completed copy using Gparted and it does say XFS for file format.

[TerryF]

I deleted the previous install to the disks by overwriting with ext4. From what you say perhaps I could have left the old one in place to be overwritten or simply deleted to start with a blank non formatted disk?

Usually this is enough, although it is better to simpley wipe the  drive using dd

I did not change the Custom setting of the screen during install or enter it , I just checked the completed copy using Gparted and it does say XFS for file format.

Looking good then.

[k_graham]

Usually this is enough, although it is better to simpley wipe the  drive using dd

Looking good then.

Except after the install of data I again am again unable to access the server from local computers.

However I can ping the computers from the server.

With the 2nd network port enabled and 1st disabled     ifconfig    shows
enp2s0     

ll /etc/sysconfig/network-scripts  does show  these 2 files  ifcfg-eno1  and route-eno1

It occurs to me I had the SME server enabled as a domain controller , I am wondering if that could be preventing the computers from seeing the server?

[k_graham]

And the output of
ll /etc/sysconfig/network-scripts

specifically you are looking for a file called
ifcfg-eno1
and
route-eno1


And have you run through the configuration again. [edit] do not enable or configure bonding at this stage
https://wiki.koozali.org/SME_Server:Documentation:Administration_Manual:Chapter5#Configuring_your_Koozali_SME_Server


And another thought, you said you can ping from your server to other devices on the network but can't ping the server from another device. Are you using only one device to test your server against? ie server -> test device, test device -> server. If not then please use the same test device to ping to and from.

Yes those 2 files are showing

[k_graham]

then you have a big issue

how did you restored your backup ?

please return the output of

Code: [Select]

ll /sbinshould be something like this

Code: [Select]

# ll /sbin
lrwxrwxrwx 1 root root 8 28 déc  2018 /sbin -> usr/sbin
also

Code: [Select]

cat /etc/e-smith-release

This is with the install done Rememberance day

 1 root root 8Nov10 19 20 /sbin usr/sbin

and

cat /etc/e-smith-release
SME Server Release 10

I am wondering if it could be as I had set it up as a Domain controller enabled ?

Also I let the router provide the DHCP rather than SME server so in cases like now where the server iis down things can still be done.

I can save the output of the files to text files such as ifconfig >/ifconfig.txt
However I have not figured out how to save those to a USB so I could then forward them to the list.

Thanks,
Kn

[sages]

I'm getting confused as to what the actual issue is.
Is it a basic network issue? Seems to have been tested with outgoing ping from server to gateway but incoming from another machine. And network bonding was enabled/disabled at some stage.
Now you are talking about domain controllers.

What exactly is the problem?
How about starting with the basics and working up from there.

Imagine that we don't know what you have or haven't done, don't know what was or wasn't configured on your previous sme 9 server, have no idea if your problem is basic network connectivity or a samba/cifs issue. You are the only eyes on your system, the people trying to help can only go by what you write on here. At the end of the day you are the one asking for help, other's are offering to help. We don't know what you know about your system nor what you have configured.

[Jean-Philippe Pialasse]

something in the config restored or something in the custom templates. 


again try

/sbin/e-smith/audittools/templates

[TerryF]

Except after the install of data I again am again unable to access the server from local computers.

So did you do a reconfigure, reboot from the console, menu item 2, Choosing the correct eth card and server mode

With the 2nd network port enabled and 1st disabled     ifconfig    shows
enp2s0     
ll /etc/sysconfig/network-scripts  does show  these 2 files  ifcfg-eno1  and route-eno1

How are you disableing port/eth card?

I would have thought there should be a  ifcfg-enp2s0 that is the eth adapter ifconfig is  showing

example, I have two eth cards, have used server-gateway in past now only server moide, so I have in /etc/sysconfig/network-scripts   ifcfg-enp2s0 and ifcfg-enp3s0, I select enp3s0 when doing the reconfigure from the console

[k_graham]

I'm getting confused as to what the actual issue is.
Is it a basic network issue? Seems to have been tested with outgoing ping from server to gateway but incoming from another machine. And network bonding was enabled/disabled at some stage.
Now you are talking about domain controllers.

What exactly is the problem?
How about starting with the basics and working up from there.

Imagine that we don't know what you have or haven't done, don't know what was or wasn't configured on your previous sme 9 server, have no idea if your problem is basic network connectivity or a samba/cifs issue. You are the only eyes on your system, the people trying to help can only go by what you write on here. At the end of the day you are the one asking for help, other's are offering to help. We don't know what you know about your system nor what you have configured.

in this last setup I had a working sme 10, and network worked, after a console restore I could ping a workstation 10.0.0.10 and gateway from the server 10.0.0.1. However when trying to ping the server from 10.0.0.1 and another computer it does not.

[TerryF]

Your restore has effed up the newtwork setup that was done during the initial sme10 install..

See what I posted above..

You have network cfg files trying to use an adapter called   ifcfg-eno1  yet ifcfg shows its enp2s0

Did you do the reconfigure from the cosole menu item2, and reboot?

[TerryF]

should also add, makeing sure to select the correct eth adapter when doing the reconfig

[TerryF]

for example this is the select local network device window when I do a reconfigure from the console on my sme10 home server, as you can see there are two adapters, its a matter of schoosing the one you have connected to your router/switch

see image

[k_graham]

for example this is the select local network device window when I do a reconfigure from the console on my sme10 home server, as you can see there are two adapters, its a matter of schoosing the one you have connected to your router/switch

see image

Yes and it works as noted that I can ping other computers from the server. But other computers are unable to ping the server. Also note both directions work before the restore of 9.2 data. After the restore it is 1 way only, even after a reconfigure. i will try to troubleshoot again this evening, suggestions welcome.

I can save command output to text file but have not figured how to save ro a USB. Perhaps to another computer so I can easily share the output.
Ifconfig.  > 10.0.0.10/???/ifconfig.txt       

Thanks Ken


Thanks
Ken

[k_graham]

something in the config restored or something in the custom templates. 


again try

/sbin/e-smith/audittools/templates

typing that at a root prompt with a return  provided 2 lines of output ending with

99SMEsite_Maker_cpwiki : Manually_Added Addition
99SMEsite_Maker_mediawiki : Manually_Added Addition

Which means ?

Thanks,
Ken Graham

[ReetP]

You need to go back to the very start here.

You are confusing the hell out of people with half answer and precious little data. That is just wasting your time, and ours, and leading to lots of frustration and no solutions.

You have GOT to be methodical and precise.

You need to prepare yourself with some tools. Notepad and pen will help. Learning what some basic commands do will help as well - if you see something you don't understand, try looking it up using the wiki here or the global interwebs. We are here to try and fix your problem, but it is isn't Kindergarten for basic Linux knowledge. You need to help yourself a bit.

So this is frustrating to see.

but have not figured how to save ro a USB

Really? You haven't tried a general web lookup or searched here? And you have bonding enabled?

Code: [Select]

tail -f /var/log/messages
Plugin USB stick

Find the device name eg

Code: [Select]

/dev/sdx

Code: [Select]

mkdir -p /media/usb
mount /dev/sdx /media/usb
Look in /media/usb

Write down steps and commands you have used as you go so we can try and repeat them. If we can't do that we can't help you. Did I mention "be precise"? Don't guess. Just do exactly we ask.

You are probably going to have to undo/remove your network bond on v9 as we suspect that may be part of your issue. Focus on just ONE card/port - if you can, remove any others. You can add other stuff back later.

Somehow/somewhere you are not reconfiguring your network correctly after the restore. Until you have got that fixed forget domain logins or anything else. Just focus on the ONE thing which is getting one network adaptor running correctly.

One way ping that you saw means that your routing/networking is probably broken. Forget dig or anthing else. Concentrate on the basic card settings. They are determined from your configurations and the network-scripts.

You need to run these on your v9 server so we can see the damage there. Make sure they are labelled as v9 data.

Code: [Select]

/sbin/e-smith/audittools/templates
/sbin/e-smith/audittools/newrpms
ifconfig
ll /etc/sysconfig/network-scripts

Code: [Select]

config show

and remove any passwords and post it somewhere like pastebin or a gist

Or run the debug report from your server-manager from the link bottom left.



On your restored v10 do the same. Label it correctly.

Code: [Select]

/sbin/e-smith/audittools/templates
/sbin/e-smith/audittools/newrpms

Code: [Select]

config show

and remove any passwords and post it somewhere like pastebin or a gist