Posts made by p4cm4n

@p4cm4n the reason ended up being that in config.class.php the page still had database_host as the old IP address
my thanks go out to sebastian who helped me out with this

@Sebastian-Roth only reason i did that was because i wasn’t on site.
sorry though, wasn’t available because someone in my company decided to do some vlan changing without letting anyone know/testing.
i’ll be available whenever probably, i’ll check in.

@Sebastian-Roth i’m more than willing to troubleshoot - didn’t want to waste your guy’s time.
considering I will be going through this exact process (maybe with less than 5 storage nodes per site however) about 12 times at the different schools i will be maintaining, i’m all for fixing.
i have opened up both SSH and web gui to the outside, and am available to be pointed in whichever direction.
further troubleshooting from me, i pulled up chrome inspector on the site locally and saw that scripts were taking a long time, along with something else which i forget at the moment.

@Wayne-Workman never before. i’ll check it out - thanks.

the makefogmobile project would help me out probably too (for little one and done labs here and there)
Thanks for that!

i’ll see if it helps my current issue. if not i’ll just reinstall them all.

@Sebastian-Roth it seems very similar to this
https://forums.fogproject.org/topic/11581/slow-webinterface-after-update-to-1-5-0

all are 1.5.4 though, only installed 5 days ago, registered 3 clients and captured 3 images.

the errors i did see seem to be with children processes. i’m not sure what it was set to before the install, but its now 5. i bumped it to 35 per george’s comments in another thread about the issue i was getting (fastcgi header)
what made it jump so low? its running on an older machine, but not that bad (core2 t8300, 2GB on debian w/xfce)

altogether though, it didnt resolve anything. i changed that, checked again and about 120 seconds or so the page pulled up. ive been continually checking both error logs and the access logs but nothing is showing in either errors now since i fixed that, put my original .fogsettings back (i tried passwords and things last night) and reran installer, and just to be sure rechecked the children processes again.

also i noticed if i just go to the apache pages, they’re instantaneous. don’t know if that helps…as soon as i clock the folder fog, is when it slows down.

edit: okay, so now i’ve noticed that if i go to a page, settings for example itll take a while to get there, but if i go to sub pages, they’re normal-ish speed. same with storage - i clicked on it to change IP’s and navigating through those pages, and saving settings were fine also.

so i changed the ip from 10.10.11.151 (dhcp) to 10.10.8.80 (static). rebooted. changed ip in .fogsettings. to new ip, ran installer. went to fogserver/fog - page loads. logged in…and nothing. just times out after a while. management page and everything beyond it all acts the same, loading…

i’ve moved them before but not these ones - these were fresh installs, the server master had 5 storage nodes below it. the one storage node i reinstalled pulls up white screen to ‘database connection unavailable’
what log/where can i find whats happening?

edit:: while i wrote this, the page pulled up to dashboard. however it still takes forever to load it, i hit refresh and its still going (3 minutes and counting)

@dylan123 it completes, and goes through its normal completion process. it looks exactly as any other machine i’ve imaged actually, boots up the same, except for throwing a few more ACPI errors after booting the kernel.
it brings up the same imaging window, 500gb HDD, 37GB in use, finishes as it should, updates database, restarts.

@JGallo slave node, do you mean a full install of fog? or a storage node install?
i can’t probably help you but i’m curious as to how you have it.

@dylan123 let me know how you figure this out - i actually did this yesterday as well and was able to capture a 37gb live usable image and transfer it over no issues. it was 500gb to 500gb HDD’s, but still finished around 7 minutes.
no idea how to shrink the partition yet but i’m sure its not too difficult i’m not a mac guru tho. also it sucks for higher versions (10.12 +) because of firmware changes.

@Sebastian-Roth
thanks! this helped me to figure out the issue, but don’t know what caused it.
it was not added to the main fog server. unfortunately i did a whole crapload of commands in sql before i checked, so i’m not sure which one was the fix. i did the command you posted above, it said unknown column password in field list.
i then ran
set password for fogstorage (cant find any matching row in user table)
grand all on fog to fog@localhost (cant find any matching row in user table)
create user fog@localhost identified by ‘password’; (query ok, 0/0.0)
create user fogstorage@localhost identified by ‘password’ (query ok, 0/0.0)
create user fogstorage@IP_OF_STORAGE_NODE identified by ‘password’ (error 1396, create user failed for <)

at this point i realized caps mattered

GRANT ALL ON fog.* TO ‘fog’@‘localhost’; (query ok, 0/0.05)
GRANT ALL ON fog.* TO ‘fogstorage’@‘localhost’; (ok, 0/0.0)
GRANT ALL ON fog.* TO ‘fogstorage’@‘IP_OF_NODE’; (query ok, 0/0.0)
SET PASSWORD FOR ‘fogstorage’@‘%’=PASSWORD(‘password’); (query ok, 1 warning)

now it works as the others were in my seperate setup.

as a test, I reinstalled a fog server in normal and three more in storage to connect to the new one. this worked fine, and I see how its supposed to work now. I have a feeling my sql is messed up from the transfer from my original to my new server.

Hi Guys,

I’ve been using this thing for months and I love it.

This question is now about storage nodes - and maybe because I screwed something up.

So I used to setup my fogserver as a debian with two NIC’s, one for WAN and one for its seperate fogserver DHCP LAN. This was a hassle. After finally messing with 66/67 settings and realizing how easy it is, I went this route.

I migrated my database, images and snapins from one VM running debian to a second machine running mint, setup as a DHCP client with pxe server. This went perfectly fine. I used the guides on here and the wiki to do this. I think I had to copy the snapins myself but ultimately ended up just using the files and re-uploading due to hashes.

So now I’m trying to expand, and have several old model laptops with 80-120-160GB HDD’s that I figured i’d throw debian/xfce on and put as storage nodes. So I installed debian base, with just system utilities and xfce, and downloaded fog, sudo, edited sudoers and ran the installer my first time as storage node.

One thing I was expecting but did not see was the database pause in the setup, when you have to go to the fogserver web console and update the schema. Not sure if it was supposed to, but if it was this is where I figured my issue lied.

So I installed it, ran everything fine, put in the fogstorage//************ into the credentials in the installer script, and all was well. Went on my fogserver and added it as storage, seemed to go fine.

The graphs all show, and list both machines - correctly. Disk usage is correct, Network throughput etc. If you click on the pie graph for fogstoragenode, it brings up the info page but says “A valid database connection could not be made”@@general in table on the right under kernel I think IIRC.

The storage node webserver brings me to a white page after redirecting to IP/fog/management/index.php?node=schema and the text says database connection unavailable.

Some troubleshooting steps I saw in forum posts and the like have me troubleshooting the database connection, so I go on storagenode and type mysql -h localhost -u fogstorage -D fog -p and then enter password, it says error 1049 unknown database fog. So I figured something was wrong there too. I did try to change the password on everything to make sure its all right ( i have that randomized sql password memorized now after typing so many times on server and storage node ) but still no luck. I tried a few other steps too but don’t remember off the top of my head.

I’m wondering if it was supposed to give me the database schema page on the storagenode. If it wasn’t I’ve re-ran the installer several times and deleted the .fogsettings file to make sure it would re-run entirely. I’ve re-ran the installer on the fogserver as well.

My hopes were to make an image for debian with fogserver in storagemode installed, capture it on fogserver, then bring a main fogserver with me to client sites. Bring several e6400’s with me, and deploy the storagenode image on them, re-run installer with new IP scheme, and use the expanded ability to send out more images per site.

so i’m in a filtered network, that i cannot specifically unfilter…but i can’t seem to get fog to install. i thought maybe it was a distro problem so i used debian and ubuntu instead of mint 19 (then 18) but same problem on all.

when it gets to adding distro it fails. then subsequently, php7.1 addons fail (like bcmath and mcrypt and all that).

so all in all the whole install fails.

whats the domain that is trying to be added or what is the repo that i can maybe just add myself? to test anyways…

scott

@wayne-workman that was basically it. worked extraordinarily well with the one. my initial setup and testing, i didn’t have control over the dhcp server so i had to do it with two. it worked at that site, but there were flaws. once deploying it, a mix of hardware issues, network issues, and ultimately permissions on things caused problems.
thanks for all you guys do though.

@wayne-workman 1.5.4.

ultimately i’ve had a crapton of problems with fog - but now I realize why. Its terrible (at least how I’ve tried to set it up) with multiple nics…and terribly easy to set it up with option 66/67. You guys can all disregard this issue - I threw a VM of it on an esx server with one virtual nic, and fog and linux are the most stable i’ve had it yet.

thanks though.

I’ve moved my fog server from a machine with 2 NIC cards, one of which was a static IP address (10.2.2.1) running dhcp, etc for booting. The other was a DHCP client using a reserved IP (172.16.x.x) to connect up to client machines for snapins and such.

I’d boot to a local network, get a 10.2.2.x address, register in fog, deploy image, reboot client machine, in my post setup install a wifi profile, install fog to the dhcp client (172.16.x.x) to receive snapins.

I’ve now migrated to using only one network card.

I changed the network card settings to DHCP from static, i re-ran the installer, which didn’t work the first time so I renamed the .fogsettings file and reinstalled. maybe this is how I broke it.

Now so far, I’ve had to change the IP in tftp and in pxe boot menu settings, which finally ended up working, but my permissions seem to be off or something. snapins register as the wrong hash, according to the client, and recreating them as a new snapin say no hash exists…I didnt know if the hash was built off of the IP, or what, but I’m stuck.

@NT_Tech product key censor?? thx tho, i’m saving your stuff for future reference too.

@sudburr

Wow.
I just was attempting to help without researching because i had this happen to me and decided…eh screw copy profile. didn’t have time to figure the mess out.

i appreciate it tremendously because what all you just posted will save me a huge headache these next couple batches of images I have to make.

clearly you’ve done your research - THANK YOU.

Hi Guys,

I’ve seen you around the forums, and like everyone always says…fog is awesome.

I got a project at my first IT job in 10 years of deploying a windows image to 300 Dell cart laptops. The laptops suck, but fog has made my job 10000000% easier.

I was wondering though, if now or in the future, a possibility of retrying a snapin if an error exists? Like for example, now I am deploying around 10 snapins to each machine - and google chrome is notorious for this in my installations, of failing with an error code resembling ‘another installation is already in progress’. this ultimately causes me to reinstall that plugin to the effected machines, but i’m lucky in that I can see the desktop, and see the icon, and make sure it is installed before I wrap it up. on things that are silent and show no output, like a management client (we use crappy centrastage) or java…if it doesn’t show as a program folder in C:\Program files (x86) or just program files…then i’ve no idea its been deployed or not. I can check - but on a host by host basis, not as a group. So if I deploy 10 pieces of software to 30 machines as a group, then 30 more, then 30 more x10…you get the idea.

Just wondering, or maybe someone as asked.