In my case, i have severals DHCP on my network and this script (i suppose) make the ipxe boot taking the wrong dhcp server and trying to boot on it. You can see here the dhcp is swapping between 172.31.43.1 and 172.31.37.252. Each time i restart the fog isc dhcp server, it took the good one but each time i reboot the client, 1/2 chance the iPxe took the wrong one because my other dhcp is responding. Is that strange ?

@Jean-Jacques-Morda Hey, this is funny, when i test with “clonezilla server”, it’s reversed. It pause on the Giga network and does not pause on the 100MB network !!! Definitively a config problem i think. More i work with network, more complex i found them. Don’t even know how those work as well with such complex procedures !

Hi,

I have several hosts on my fog server.

Naming are : ClassroomTypeNumber, for example :

A111PC01 for the PC 01 of the A111 classroom
A111PCPR for the teacher PC of the A111 classroom

When i lauch a new search on hosts with “pcpr” string, it display me

a111pcpr
b110pcpr
b010pcpr

etc…

but also :

c005pc01
c005pc02
c005pc03
…
c005pc17
C005pcpr

but c005pc01 to c005pc17 doest not contain “pcpr” string.

Hi guys,

I have another (yes again) problem with a computer.

Cloning this one with “clonezilla” using partclone is working in uploading as well as downloading.

With Fog :
Uploading is ok.
Downloading in mode “multi partitions single disk not resizeable - Everithing” display me : Error reading logicial partitions. List may be truncated!

If i try to restore primary only, it’s working.
If i try to restore logicial : it does not display error then done process without restoring anything.

Partitioning is :
d1p1 primary
d1p2 extended
d1p5 logicial

I suppose that i have a problem with my partitions but why partclone and clonezilla are working and not fog ? What fog is doing when testing to read “logicial partition” without failing uploading it ?

Hehe, don’t want to abuse but is a “per group” --max-bitrate option is possible ? Some computers groups are connected to a part of the network that need to be limited with --max-bitrate and some others not. Not very important but it will avoid to change master node setting when we launch cloning with another group. I suppose it’s a little bit more complex to implement

@Wayne-Workman Hmmm… every image i upload yes but in the 100MB network plugs. If i connect to Giga Network, upload does not pause. Strange. Peharps a link with network cache configuration ?

@Tom-Elliott If i change of computer to connect to the server, it’s the same. It’s not a cache or browser problem. Display updating informations but they are stuck at a certain point. Client side seems to send always the same information but continue to upload or download.

Others tests these days says that it do it in uploading/downloading, unicast and multicast. All tasks can freeze but continue on client. Web display is refreshing but as i said with the same informations. If i change browser or use another computer to connect tasks, same information is displaying with no progress. I think it came from client side but not sure. How are progress informations sent to tasks display ?

SO…

After moving partition to reduce space between mbr and first partition, i success to restore my image. By the way, dev-branch add the constraint to reduce max mbr size… so, i suppose problem is solved.

Hi,

Somebody know why, when uploading images only, partclone upload 1%, 2%, less or more then pause for 10 seconds or 2 minutes then continue 3%, 4%, then pause again for 30 seconds then continue 5%, then pause for 1 minute, then continue… then pause, then continue, then pause, then continue… seem to be randomly and, finally, it took twice the time to upload an image in front of clonezilla that does not have the problem. Frustrating. Timer is continuing while blocks stuck… then restart… then stuck… then restart… then stuck…

@ch3i yes but if we close the browser and/or disconnect, it’s the same. We can even change computer to connect thru the web interface, it’s stuck. Maybe it’s an “overflow” or something like that between partclone status and web interface, don’t know.

@Wayne-Workman yes, it detect the max theoretical speed… like speed on road. On road, sometime, we have to slow down to avoid accidents, if not, cars does not arrive to destination.

@cspence Thank you, i’ll wait I’m testing uploading an image with moved first partition to see if my problem was due to the size of the crap before the first partition… result in two hours… i would like to discover why this one does not restore !

In fact, i didn’t noticed before but i have a warning unploading in one disk resizeable mode :

After resize test … /usr/share/fog/lib/funcs.sh: line 167: warning: here-document at line 165 delimited by end-of-file (wanted ‘EOFNTFS’)

@Tom-Elliott d1.mbr : 103424K and it’s not a Dell but an HP. Honestly, i don’t remember wich type of install. I’ll try to “move” the main partition to avoid the problem i think (if i can…). But it’s interesting to test if the first partition of a disk begins a 10Mb, the mbr will be 10mb…

Yes i noticed that. When we have only name/password, it’s ok. It was just a suggestion Prepopulate is a good idea when we have a lot of users to add with a lot of fields to fill with same informations. My mistake came from I usually use “back” button on some web interface to add several users with keeping fields filled with my information. The habit…

Yes, i’ve deleted the recovery partition because i never use those recoveries plenty of craps but i don’t think the is a link with the mbr problem. My others computers are working well but partitioning scheme is not the same.

I found clues … :

On working images, first partition is starting at 32kb … but with this one, coincidence, partition starts at 106mb … I think the mbr procedure is backuping all before the start of the first partition, not only mbr.

my needs is 80m so i set it to 80m and all is ok now for my network !!!

As i’m in charge of several users, when i add one, whatever the program is, i’m waiting for, after clicking on “add”, to directly go to add another one instead of go to edit the user i’ve just added. Peharps two buttons :

create user and edit
create user and add another one

or something like that ?

Just a suggestion. I’m thinking about this because, some days ago, i’ve add several users before seeing that i’ve edited the same again and again… :bowtie:

SVN 3710 (but others as well) :

In task list, when multicast download, actives tasks stuck around 70% only in the web display when downloading big images (more than 150gb), even if i log out / close my browser and whatever the browser. Tasks dissapear when finished but does not go further than around 70% for 150gb images… but multicasting is still working and all is ok at the end of the process. It’s just the active task display that is bugging. Seems to work with the old fog 1.2 but was it luck ?

Yes, that was this option help me to find out how to deactivate it… but, my request was to avoid deactivate for all, just for computers that are bugging with that. All of mine this case but i certainly will have the case that some will works and others not. A by group WOL activation should be really cool.