Posts made by Wayne Workman

@benc said in Seems like you are trying to restore to an empty disk. Be aware this will most probably cause trouble.:

Would you suggest doing anything to Ubuntu as far as upgrading or updating?

The fog installer ensures the packages it uses gets updated, so you don’t need to worry about updates. But - if you want an updated OS, you could absolutely update the system (after you snapshot). The one-liner I use for updating Ubuntu and Debian is:
apt-get update;apt-get -y dist-upgrade;apt-get -y autoclean;apt-get -y autoremove;reboot
Updating doesn’t hurt anything, and it’s good practice, as long as you have a healthy snapshot to go back to should something go wrong.

@CJR_ie What version of Fog are you using?

In addition to what Quazz has suggested, you could also take a snapshot of your VM (for backup), and then upgrade to the working branch. There’s a lot of fixes in there. Here’s instructions: https://wiki.fogproject.org/wiki/index.php?title=Getting_FOG If this doesn’t work out for you, you can always revert to the snapshot and it’ll be like it never was done.

@benc Is your fog server run in a VM or on bare metal? I ask because there are options to try if it’s in a VM that otherwise are not very easily possible.

@sapeurorca Try to reset encryption for that host, try to re-install the fog client on that host.

@Tom-Elliott Store them in the database and have a daily cron job that cleans out logs older than N days (user defined). If FOG were to switch to MariaDB 10.3, the TokuDB engine would be available for use. TokuDB has in-engine compression. The compression savings is pretty staggering compared to InnoDB. TokuDB does not support foreign constraints - and because the FOG db doesn’t use them, we are A-OK there.

@george1421 said in Bulk editing hosts:

If you are going to mass update the database like this, stop apache and php-fpm first, then run your update, and finally restart php-fpm and then apache. Both php-fpm and apache cache database values for faster access. You are changing the underlying database but its not seen in the UI because of the cache.

#wiki worthy.

I have tests that automatically test Ubuntu 18.04 automatically every day. Links in my signature.
I’ve been told by @Sebastian-Roth that there is a big repository difference between Ubuntu 18.04 LIVE and non-live, namely the universe repository is present in one, missing in another.

@andreiv Have you seen this article yet? https://wiki.fogproject.org/wiki/index.php?title=Troubleshoot_Downloading_-_Multicast

Just adding a note here. Typically when the TFTP server address is requested at a prompt, this in the past has several times been due to multiple DHCP servers replying with different option 066 addresses. Typically, a rogue DHCP server or two redundant DHCP servers that are simply mis-configured.

@Boyan-Biandov said in upgraded to Ubuntu 18.04.1 LTS:

Is it really necessary to copy the images to a folder, other than /images on the new fog server and then once they are copied there “rearrange” them by moving them again into the final destination of /images (assuming fog defaults)?

Yes, because FOG’s NFS configuration does not allow writing to /images, it’s a read-only share. /images/dev is read+write. I kept the walkthrough as simple as possible, leveraging things already setup on the servers. If you wanted to use something like SCP, you could put the images directly where they need to go.

@bsalogga This article was made for you: https://wiki.fogproject.org/wiki/index.php?title=Migrate_FOG

You could also disable the nodes you want to delay replication for via Storage Management.

@Sebastian-Roth is right, the logs will tell us more. Also, Tom did fix the issue of the replication sleep cycle maybe two years ago. if lftp processes are still running, no more get started. They have timeouts so no fears of infinite hanging processes.

Also I’m just going to say that most replication issues I’ve seen people have are normally due to configuration problems. Like a bad fogstorage username & pass, or MySQL remote access being disabled, or firewall on the OS blocking the needed ports, or bad FTP username/pass for the storage node.

I appreciate what you do for FOG. If you need VM space, you know you can give me a hollar.

@BenR We need a section from the fog log file to help you. It’s here: C:\fog.log The errors in there are the actual windows errors returned from windows. There will be an error code. If you post a big chunk of the log (including error), I will take that and go look it up here and tell you what’s wrong.

@mstumpo said in How do I configure the Fog server to use WIndows Server for image repository?:

That’s my only option? I can’t mount a Windows share in Linux and point to that?

There has been limited success in using an NFS share in Windows. @george1421 has a walkthrough here and here. Another guy posted a powershell script that does all of it but I can’t seem to find it. This is definitely not something most fog users do, it’s not a regular practice. If it breaks, you’d be hard-pressed to get help with it.

Another option is ISCSI. If you had an ISCSI target, you could mount that on the fog server, no problems as the OS treats it as a local disk.

But, as far as just making a regular CIFS/SMB share on windows and then mounting that on the fog server - it won’t work for imaging because the Linux does not allow exporting a directory that is already remote.

Also, there is the Hyper-V option, where you can just setup a storage node, or just a whole new fog server - you can move everything over to the new server by following this tutorial: https://wiki.fogproject.org/wiki/index.php?title=Migrate_FOG Having FOG virtualized provides beneficial features like snapshots & easy migration to new hardware.

@mronh On the storage node, check the free space. use this: df -h and look for partitions that are 99% or 100% used.

@mstumpo Run your FOG Server in Hyper-V on that windows server, and give it lots of space, make it thin-provisioned.

@asbenavides

• What version of FOG?
• What OS & version is your FOG server?
• Have you done this before?
• Did you run the storage node installer in storage node mode?
• Did you input the credentials correctly during the storage node installation?