Latest FOG 0.33b
-
Pxe boot menu configuration page seems to work fine now thanks.
I checked on 0.32, it wasn’t resizable, I just hadn’t noticed the change to box at 0.33bThanks again.
-
r1200 released.
Adds delay and display of what the partitions are set after mbr/gpt generates the partition tables. Adds image name lookup on task search. Fixes the padding issue on the progress bar. Updates the TomElliott.config to 3.13.2 setup.
-
r1201 and r1202 released.
r1201 uses partclone.imager for raw image formats in mps and mpa imaging. This way, the partclone.restore actually works if an image had to be dd copied.
r1202 adds Storage Node elements to the Location Plugin stuff. Adds the same to the GroupManagementPage, HostManagementPage and LocationManagementPage. I have not added the host to the location stuff yet, but that will be in due time. Word of caution, if you have the location plugin enabled and you register a host, it will be required to enter a location for that host. If you don’t it will cause issues for the moment. I haven’t entered this in yet. So for the mean time, just be cognizant of this.
-
r1203 released to add some fixes to the Location Plugin.
Right now, my logic is:
Location To Group, select all storage nodes within that group. May not work right away as it checks for all storage nodes in the system, will need to add checking against the groups.
-
r1204 and 1205 released.
More modifications to Location stuffs. r1204 moved the location element to above host/group kernel items. Fixed typo for the add host location portion. Adds checking if the node is of a different storage group, it will change the group id to that of the node’s so things still work.
r1205 fixed issue in generating a new location. Needed to verify if storagenode was being set then check the group id and change it as necessary. This is only valid on the Adding a location. Need to add element when editing to only get the storage nodes within the specified group.
-
I Hey,
I updated to revision 1205 and tested the deploy of the (new uploaded) image again. The init.gz has been replaced. Fog shows me
Listing of recognized partitions:
/dev/sda1 (C: partition), /dev/sda2 (extended partition, contains /dev/sda5 and /dev/sda6)
Partlcone starts and then it exits with “error exit” for both partitions and fog shows me again "no such file or directory /dev/sda1.
Starting the task again, partclone and fogs then works fine.Do you think it would help to reinstall ubuntu?
-
No,
I’ve been trying to replicate the issues you’re seeing as I’m just not seeing it. That’s part of why I added the output of what partitions it thinks it has. The partitions you’re seeing are what’s created directly after the MBR or GPT tables are generated. My question about this issues though, is the BIOS set to SecureBoot and UEFI? And the OS is installed as MBR? I might think about, basically, righting the MBR/GUID tables twice and still outputting the partitions. This might help out with what we’re seeing.
-
I use VirtualBox for testing and there I have no UEFI activated. The only thing that I do is, I clean the disk with gparted, removing all partitions.
-
Well,
I added sgdisk -Z which removes all partitions (GPT or MBR) and mbr data before “installing” the new partition tables.
-
I think I found the issue. I’m a dumb*** lol.
Basically, I used copy paste which by itself is fine. Except the line that’s supposed to remove the partitioning tables for MPS tasks was:
[code]sgdisk -Z $disk >/dev/null[/code]
The problem with this is that it works perfectly with MPA format because I have a variable called Disk. in MPS it only needs to worry about the current hard disk.
So I’m uploading right now.
r1206 should fix the mbr issue. Maybe it’ll work properly now? Please test and let me know. Sorry about that over sight. It’s only piece of code to look through 400 lines so I hope you’ll understand my mistake there.
-
r1207 released.
I was finally able to replicate the error exit issue described above.
My systems would work, as far as I could tell, because the partition tables weren’t truly changed on my systems.
So I completely wiped my drive and tried again, finally able to replicate the issue. It looks like the first iteration the partition table doesn’t actually get written.
So the interim fix is to add the fdisk command to simply write the tables to the disk.
-
hey question for you tom is there any way to inject items into the image a addon-image if you will?
-
Not easily, though I’ve heard of people doing this kind of thing with partclone images, I’ve never done this myself. Theoretically it is possible with partclone, but I don’t know how to do it myself.
-
I will look into this I think it would be a great addition to be able to do this. it would avoid having to take an image before sysprepring or what not and it would just allow us to add to the image.
-
r1208, r1209, and r1210 released.
r1208 just made correction to storage node identification. If there isn’t one set on location, it tells you not set.
r1209 fixed an issue in the host management page with location not being able to be adjusted.
r1210 Adds updating of inventory on all download tasks. This way, if you manually added a host, it will update the inventory with the systems information. Yay, better inventory management.
-
Just a little update on the Location Plugin.
I’m told it works properly. One of the largest, and I think influential changes, of Lee’s original Location Patch is that it doesn’t have to be tied to a specific node. In the original patch, you had to tie the location with a node. That node would then be the goto place for that particular system.
Under the new logic method, it’s a plugin. This means there’s two new tables created.
The first table within the schema change is the location table. This table is what holds the information on a specific location. The methodology of this is you name the location and, at the very least, assign a storage group to that location. You may also assign a specific node to the location. It will update the Storage Group to whichever group the node belongs in. I’m sure there’s a few kinks to iron out, but it seems to work from what I’ve been able to test.
The second table is locationAssoc. This is how (change from the patch method) it associates the location to a host. You will, most likely, need to assign the location to the host if you’ve enabled this plugin for all new registration’s after the location has been enabled.
What this allows?
The biggest difference is that this allows you to use Groupings AS WELL AS specific nodes for where you want the client to get it’s images. If this person moves around a lot between locations, you just make the change to that host and all will work fine. This makes for a centralized place to manage all of the hosts under your control.
I hope all will enjoy this being tied in as a “plugin” to the FOG system.
Many many thanks to Lee Rowlett who developed this originally and the assistance with getting many of the kinks ironed out and the details of how to go about getting this implemented. I admit I added new features that were unexpected initially, but I think it allows for more dynamics and adaptability within any organization.
-
Tom you and lee rock I will give a go gotta get my server back up today now as you know I just moved. But I will let you know how I do with it. will there be updated instructions for any one new who may need this?
-
The instruction set is much simpler now.
Just enable the plugins on the fog server.
Go to activate plugins.
Click on Location.
Go to installed plugins.
Click on Location.
Click install plugin.
Then click home. You’ll see the location Icon appear.
Click on the location icon and setup your location as you normally would.
Go to your hosts and set up your location.
I also added the element (which may have been there already) to Add all hosts within a group to a location as well.
That’s it.
It looks like a lot when you read it, but really it’s only seconds of work.
-
sounds easy will let you know
-
Hello
To start I apologize for my English.
Fog is a very good job and thank you for this.
is it normal not to see the password FOG_STORAGENODE_MYSQLPASS in fog setting?
When i add a storage node is asking me for this paswword?And just for the report:
When i go to the service management and I selected any service the graphics of the page disappearsThank
