Latest Development FOG
-
[quote=“Tom Elliott, post: 38512, member: 7271”]When we decided to move to ipxe, we knew this would come up and was a potential candidate to register systems. The reason we didn’t is while iPXE and even good old pxe can recognize hardware and systems, the OS layer (bzImage/init.xz) does not necessarily have the same level of detection. At the BIOS level (where ipxe and pxe operate) it has access to the bios information which is what’s presented. Part of why we see some issues with ipxe booting altogether is because drivers are a whole separate animal. The OS layer uses more information to determine the best flow for the workload that the system is being presented.
Simply put, the reason we don’t register systems from iPXE is simply to ensure a system is compatible the whole way through. If a system cannot be registered in OS layer, you can’t image. This is why we register at the OS level rather than through iPXE.[/quote]
Tom, I’m sorry, maybe i mislead you with my question, let me try to clarify my question.
when we iPXE boot the system, we are presented with the menu of options. if the system was previously registered its green indicating FOG recognizes the system and has it in its database of systems, if its red it shows that this system needs to be registered. Love this feature makes the process of knowing if its associated much faster for the guy on the go.
What I am however asking is more simplistic. I’m asking if when we get to the main menu in iPXE where it provides your options to run memtest, qick image, ETC is it possible to add the option to create (manually via typing in a name) the place holder for the new image about to be sent to the fog server? and then if a name is provided offering up the opportunity to then upload a new image from that machine to store as a deploy able image for fog to use for future deployments.?Im looking to find a way to streamline image creation for the various pc’s we seem to be acquiring from IBM/Lenovo. would like to just be able to walk up boot the pc and then register the pc and then create an image name for it /register it to that pc and then pull the image from said new pc to store in fog to then deploy to units as needed.
This is a similar process to the already provided option of imaging the unit in place when its registered.
I hope i have made my question easier and more to the point. again, thanks for all the hard work this is a life saver for our facility and its been a fun experience to play with.
-
[quote=“Tom Elliott, post: 38519, member: 7271”]I’m not trying to shun suggestions or ideas, it it’s for things such as this I created the menu customization as dynamic as I could.
Plus, on an upside, if you create the items and make it successful, report back how, we have a kind of documentation. If I just do everything that people ask, why even have the customization stuff available?[/quote]
Sorry Tom, Just saw this, I wish I had more knowledge about how to customize these options I would gladly dive in and learn about it, i was only asking if it was something that was simply added just like changing the text . I am by no means a programmer and wish I had those skills but i try my best to identify ways to better support our facility and was curious about this process.
-
Tom,
In the latest SVN, what php file is storing the settings for custom ipxe boot menu ? When you add a custom menu item basically … what file stores that . Reading through the forums its under advanced but im going through all the code and im not coming across it.
-
Its pulled from the database no files.
-
Any documentation on adding the new custom IPXE Boot menu ? Just want to link my custom bootmenu item to an .iso.
-
there is no documentation, technically the feature isn’t released yet.
the method/syntax is very similar to what is done on the advanced boot menu, however -
Nothing directly but looking at the regular entries can help give an idea of what to do
-
So under paremeters whats the easiest way for me to link the .iso in the tftboot folder ?
-
you don’t need to link to the tftpboot, nor would I recommend it.
Linking an ISO in the tftpboot would mean you’re attempting to load the iso through tftp protocols. This tftp protocol is limited to 10Mbps (around) transmit rates.
-
personally, i create an /iso directory in my web root, and pull my disk images over http
-
SVN 2518 released.
I don’t know if it will work, but rather than pulling the entire query into a single variable, I’m using prepared statements which are supposedly unbuffered. Hopefully this will fix the issue of too much memory usage, though I’m suspecting it really won’t help at all. Either way, things seem to work on my side. I don’t have a large enough data set to have issues with this though either.
This release also moves all of the GUI stuff that referenced the location plugin in the source code to purely be based on the plugin. This way the plugin is a real plugin, not something that’s kind of a mix between core and plugin.
-
SVN 2519 released.
I had to revert the method’s of the mysql stuff back to the original as prepared statements won’t run for some reason with the prepared statements.
Hopefully we can figure something out.
-
SVN 2526 released.
Yes, quite a few. Basically it just adds the functionality to email yourself when an imaging task completes. Thanks Lee Rowlett.
I’ve heard of others requesting this functionality as well so hopefully you find it useful.
It defaults the email to disabled, but it’s in FOG Configuration->FOG Settings->FOG Email Settings.
You’ll find:
FOG_EMAIL_ACTION, checked means enabled, unchecked means disabled.
FOG_EMAIL_BINARY, used to set the binary path, usually sendmail.
FOG_EMAIL_ADDRESS, specify your emails in the form [email]email1@domain.com[/email],[email]email2@papa.com[/email]. Separate emails by comma basically.
FOG_FROM_EMAIL, specifies who the email is coming from. ${server-name} is just a holder to get the node. So noreply@${server-name}.com will translate (if node’s name is weirdo) into [email]noreply@weirdo.com[/email] You can also specify in the form ${server-name}@domain, etc…You’ll also find:
FOG_MEMORY_LIMIT,sets the memory limits of GUI directly from FOG rather than having you all edit and restart web server. It will always default to 128, even if you set below or the value is not numeric. Higher values are kept. -
Love the email addition, I currently have a line in my post deployment script that calls blat and sends an email before the shutdown command.
Thanks Tom and Lee
-
SVN 2535 release.
Snapin Replication works properly, in that it will only copy over relevant snapins, not the entire snapin directory even if one of the snapins is only intended. Also made it so it works similar to images. Meaning snapins can be assigned to multiple storage groups and the data will be transferred between those groups. The node master of that group will then distribute to the rest of the nodes in it’s grouping.
Cleaned up a few installer bits as well and fixed the ubuntu installer so SnapinReplicator server gets started.
-
Tom, I just attempted to install rev 2542 on Ubuntu 14.04 (I know) in a new vm and it failed at setting up and starting MySql. I realize 14.04 isn’t officially supported but I haven’t had this issue before - I hadn’t done anything other than installing subversion prior to the install. I’ll try a supported version and see how it goes
-
[quote=“ArchFan, post: 38593, member: 19266”]Tom, I just attempted to install rev 2542 on Ubuntu 14.04 (I know) in a new vm and it failed at setting up and starting MySql. I realize 14.04 isn’t officially supported but I haven’t had this issue before - I hadn’t done anything other than installing subversion prior to the install. I’ll try a supported version and see how it goes[/quote]
All good, just let me know how it goes.
I haven’t seen any issues on my 14.04 box though so maybe something else went ary?
-
[quote=“Tom Elliott, post: 38595, member: 7271”]All good, just let me know how it goes.
I haven’t seen any issues on my 14.04 box though so maybe something else went ary?[/quote]
Just tried 13.10 and rev 2544 - same issue. I don’t think I’m doing anything differently than I have in the past, but it’s been a long weekend
-
When you’re installing the OS, are you installing with LAMP?
-
[quote=“Tom Elliott, post: 38597, member: 7271”]When you’re installing the OS, are you installing with LAMP?[/quote]
I’ve always used the minimal ISO, and just chose ‘manual package selection’ and FOG seemed to install its dependencies during the install - has this changed?
