FOG Compatibility Test Failed

[george1421]

I agree here, if your fog server is behind a proxy server (i.e. you require to use a proxy server for internet access) there are some changes you need to make to allow this.

[techlover28]

@george1421 said:

I agree here, if your fog server is behind a proxy server (i.e. you require to use a proxy server for internet access) there are some changes you need to make to allow this.

I am connected to the internet directly.

[george1421]

@techlover28 Sorry I should have refreshed before I posted. It almost appears that your fog server can’t resolve IP addresses according to the error. Did your network team just directly connect you or has it alway been this way? It appears that dns resolution is not setup??

[Sebastian Roth]

Well then try (test-)downloading the kernel by hand and see what happens: curl -ko "/tmp/bzImage" https://fogproject.org/kernels/bzImage

[george1421]

@Sebastian-Roth said:

Well then try (test-)downloading the kernel by hand and see what happens: curl -ko "/tmp/bzImage" https://fogproject.org/kernels/bzImage

Actually this is what I was going to have the OP do to update the kernel without having to upgrade to the trunk (even though the trunk is a good choice)

[techlover28]

@george1421 said:

@techlover28 Sorry I should have refreshed before I posted. It almost appears that your fog server can’t resolve IP addresses according to the error. Did your network team just directly connect you or has it alway been this way? It appears that dns resolution is not setup??

No worries. I have an internal IP assigned to the server with a default gateway IP. It has always been this way. I can directly download/install the RPMs from atomic yum repositories. I am able to get the IP when I try to ping google or fogproject.org.

ping google.com
PING google.com (216.58.197.78) 56(84) bytes of data.
64 bytes from maa03s21-in-f14.1e100.net (216.58.197.78): icmp_seq=1 ttl=57 time=42.3 ms
^C
— google.com ping statistics —
1 packets transmitted, 1 received, 0% packet loss, time 976ms
rtt min/avg/max/mdev = 42.327/42.327/42.327/0.000 ms

ping forums.fogproject.org
PING fogproject.org (162.213.199.177) 56(84) bytes of data.
^C
— fogproject.org ping statistics —
7 packets transmitted, 0 received, 100% packet loss, time 6051ms

[george1421]

@techlover28 from the command prompt on the fog server can you successfully execute the curl command Sebastian posted below? That will test the innerworkings of the download processes. Its possible that pings are allowed but not downloads in your company firewall. The curl command will test how well the downloads work.

BTW: The target of that curl download command IS the latest 64 bit kernel. All you need to do is rename your current kernel and then move this one into place to upgrade to the latest kernel without upgrading FOG.

[techlover28]

@george1421 said:

Actually this is what I was going to have the OP do to update the kernel without having to upgrade to the trunk (even though the trunk is a good choice)

@Sebastian-Roth said:

Well then try (test-)downloading the kernel by hand and see what happens: curl -ko "/tmp/bzImage" https://fogproject.org/kernels/bzImage

Does above mean similar to below post?

@george1421 said:

@techlover28 OK then, If you navigate to your web servers base directory (usually /var/www/html or /var/html) and then go to <www_base_dir>/fog/service/ipxe/ and rename bzImage and bzImage32 to bzImage.old and bzImage32.old

Then run the following commands to download the latest kernels
wget https://fogproject.org/kernels/bzImage
wget https://fogproject.org/kernels/bzImage32

From there reboot your target device to load in the newest kernels. If something breaks with this target or any other ones, just roll back the changes you made to load the old kernels.

If yes, then it didn’t work last time when I tried. I can try again though.

@techlover28 said:

I have tried it (newer version of kernel) today. It says “Network - Fail” along with “Disk - Fail”. Older version of bzImage works with Network therefore as you mentioned, I did the rollback.

[george1421]

@techlover28 Yes that is the same. That is sad the newer kernels work not good. Maybe if you make a nice donation to the @Developers they would add your network driver and raid card to the FOG boot kernels

It looks like you may have to build your own kernel to get these specific drivers to work

[techlover28]

[[modules:composer.user_said_in, @george1421, FOG Compatibility Test Failed]]

@techlover28 Yes that is the same. That is sad the newer kernels work not good.

Ah! yes, that is sad.

Maybe if you make a nice donation to the @Developers they would add your network driver and raid card to the FOG boot kernels

I earn $12000 (minus income tax) an year, not ashamed to tell it . Giving the fact that developers are so much helping and friendly, I was already thinking about giving a donation (not a nice amount but good percentage of my monthly income). Not that I am generous, but thankful for the help and support I have been getting instantly.

It looks like you may have to build your own kernel to get these specific drivers to work

I will try it and hope it goes well. If it doesn’t work, I may have to move to some other solution as I am under pressure to finish this task soon.

[george1421]

@techlover28 Right now unless the @Developers are willing to add them to the official build, then yes building your own kernel version is the only option. Its not that hard it just take some time to setup the environment. You can get the drivers right from the intel site and then just build them in.

[Sebastian Roth]

Well wget and curl do kind of the same thing but they are still different tools. Have you tried using the command I posted? Because curl is used in the installer script. And I thought it would be a good idea to test exactly what is going on when you run the installer (which failed when downloading kernel/init/fog client).

Adding RAID drivers which are not included in the official kernel source is kind of a huge effort. We would need to keep those updated and tested. How are we going to do this not having that hardware at home.
On the other hand you can just try following the kernel compile wiki articles to build your own custom kernel. We are more than happy to help if you run into a problem with this.

The NIC is a bit of a mystery to me. Should be I350-AM4 (quad port) if I got this correct from the mainboard manual. Those should be supported by driver ‘igb’ in old and new kernels! http://cateee.net/lkddb/web-lkddb/IGB.html

[techlover28]

@Sebastian-Roth said:

Well wget and curl do kind of the same thing but they are still different tools. Have you tried using the command I posted? Because curl is used in the installer script. And I thought it would be a good idea to test exactly what is going on when you run the installer (which failed when downloading kernel/init/fog client).

Yes, I tried the command. It successfully downloaded a bzImage file under /tmp directory with size 6883968 bytes.
The download kernel part failed on the 2nd attempt, first attempt was successful which later got stuck at “Backing up database”.

Adding RAID drivers which are not included in the official kernel source is kind of a huge effort. We would need to keep those updated and tested. How are we going to do this not having that hardware at home.
On the other hand you can just try following the kernel compile wiki articles to build your own custom kernel. We are more than happy to help if you run into a problem with this.

I will surely give it a try, starting Monday.

The NIC is a bit of a mystery to me. Should be I350-AM4 (quad port) if I got this correct from the mainboard manual. Those should be supported by driver ‘igb’ in old and new kernels! http://cateee.net/lkddb/web-lkddb/IGB.html

Yes, you are correct. Please find more information below:

– ethtool -i eth0

driver: igb
version: 3.2.10-k
firmware-version: 1.4-8
bus-info: 0000:04:00.0
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes

[JJ Fullmer]

I had a problem with an older raid controller not being recognized by the centos/RHEL 7 kernel just yesterday. Would this happen to be an older hp smart array controller? Like 2008 or older?
You could try adding these host kernel arguments (in the gui for the host, register it manually if you haven’t already)

hpsa.hpsa_allow_any=1 hpsa.hpsa_simple_mode=1

That might do the trick. These older array controllers are no longer supported by the RHEL kernels, this is where I got that info…
http://serverfault.com/questions/611182/centos-7-x64-and-hp-proliant-dl360-g5-scsi-controller-compatibility

[techlover28]

hi,

I rebuild the FOG server today and was able to move further on upgrade. This time it got stuck here:

====================================================================================
Downloading inits, kernels, and the fog client…OK
Comparing checksums of kernels and inits…Done
Enabling apache and fpm services on boot…OK
Creating SSL CA…OK
Creating SSL Private Key…OK
Creating SSL Certificate…OK
Creating auth pub key and cert…OK
Resetting SSL Permissions…OK
Setting up SSL FOG Server…OK
Restarting Apache2 for fog vhost…OK
Changing permissions on apache log files…OK
Backing up database…OK

You still need to install/update your database schema.
This can be done by opening a web browser and going to:

http://192.168.1.1/fog/management

Press [Enter] key when database is updated/installed.

Setting up storage…OK
Setting up and starting DHCP Server…Failed!

So I reconfigured the DHCP server and restarted its service manually with success.

Then I commented entire “installfog.sh” script for Trunk except last few lines where it executes all the functions like “configureDHCP” etc.

I executed “installfog.sh” again and it shows me below (everytime below is shown now with the commented “installfog.sh” script):

====================================================================================

• Setting up storage…OK
• Setting up and starting DHCP Server…Skipped
• Setting up and starting TFTP and PXE Servers…/lib/common/functions.sh: line 432: : No such file or directory
  OK
• Setting up and starting VSFTP Server…/lib/common/functions.sh: line 373: : No such file or directory
  OK
• Setting up FOG Snapins…OK
• Setting up UDPCast…OK
• Configuring UDPCast…OK
• Building UDPCast…OK
• Installing UDPCast…OK
• Installing FOG System Scripts…Failed!
  ====================================================================================

[Tom Elliott]

@techlover28 you commented the lines in installfog.sh that sources the config.sh files as well.

[techlover28]

@Tom-Elliott said:

@techlover28 you commented the lines in installfog.sh that sources the config.sh files as well.

No, sorry forgot to tell. I started commenting the file after two source files (config.sh and functions.sh) lines.

[Tom Elliott]

@techlover28 please use the file provided as is. I don’t know what is or isn’t included in all the comments and what not. The only guess I have is from the before commenting part in which I’m guessing you want fog to be the dhcp server?

[techlover28]

@Tom-Elliott said:

@techlover28 please use the file provided as is. I don’t know what is or isn’t included in all the comments and what not. The only guess I have is from the before commenting part in which I’m guessing you want fog to be the dhcp server?

Yes, I want fog to be dhcp.

I have attached the commented installfog.sh file (trunk). As the upgrade failed at the point where it was setting up DHCP, so just thought if commenting some part of script could avoid complete re-installation.

0_1454379388299_installfog.sh

[techlover28]

@Tom-Elliott I re-tried installing Trunk from scratch after removing-reinstalling DHCP and VSFTPD RPMs.

It got stuck at same place again (Setting up and starting DHCP Server…Failed!).

Attaching logs for reference. 1_1454387019861_foginstall.log 0_1454387019861_fog_error_6134.log

It is not able to start the DHCP service, therefore I manually configured it and re-executed the installfog.sh script. Script overwrites the configuration and fails.

My manual configuration which let DHCP service to start:

   subnet 192.168.1.0 netmask 255.255.255.0 {
   # option subnet-mask ;
    range dynamic-bootp 192.168.1.10 192.168.1.100;
    default-lease-time 21600;
    max-lease-time 43200;
    #       option domain-name-servers      x.x.x.x; 
    #       option routers      x.x.x.x;
    filename "undionly.kpxe";
    }

Script configuration which does NOT let DHCP service to start:

    subnet 0.0.0.0 netmask  {
    option subnet-mask ;
    range dynamic-bootp 0.0.0.253 192.168.1.254;
    default-lease-time 21600;
    max-lease-time 43200;
    #       option domain-name-servers      x.x.x.x; 
    #       option routers      x.x.x.x;
    filename "undionly.kpxe";
    }