Unable to locate image store

[iekozz]

Hi all,

With all your help, I finally managed to install the fog server and put all my images on it. However, when I boot to the pxe menu and want to deploy a image, I get the following error: Unable to locate the image store

I double checked the following:

• The default storage is on and enabled
• the images are located on my disk on /images
• Every menu item I can see is linked to /images
• There is a image about 10gb in size

I have no idea what to do. Will post the error log tommorow when I get back to the shop, but I will take any suggestions now

Thanks!

[Wayne Workman]

@iekozz See these:
https://forums.fogproject.org/topic/10238/fog-1-4-2-unable-to-locate-image-store-bin-fog-download/8
https://forums.fogproject.org/topic/9187/unable-to-locate-image-store-bin-fog-download/7

[Sebastian Roth]

@iekozz Probably the .mntcheck file missing. Run ls -alR /images and post output here.

[iekozz]

@Sebastian-Roth

I get this as output:

[fogserver@localhost ~]$ ls aLR /images
ls: cannot access aLR: No such file or directory
/images:
dev  postdownloadscripts  Win10uefi
[fogserver@localhost ~]$

The dev postdownloadscripts Win10uefi line is in green

I have captured the image first (obviously) and then went to deploy it via deploy image in the pxe menu.

@Wayne-Workman said in Unable to locate image store:

@iekozz See these:
https://forums.fogproject.org/topic/10238/fog-1-4-2-unable-to-locate-image-store-bin-fog-download/8
https://forums.fogproject.org/topic/9187/unable-to-locate-image-store-bin-fog-download/7

Yes, seen those, but neither seems to be able to solve my problem, did everything that was suggested there.

[Sebastian Roth]

@iekozz Please run the command as suggested, you missed a hyphen…

[iekozz]

@Sebastian-Roth Ah, you’re right, whoops. Here is the output:

[fogserver@localhost ~]$ ls -aLR /images
/images:
.  ..  dev  .mntcheck  postdownloadscripts  Win10uefi

/images/dev:
.  ..  0800271fe49c  080027aff8e3  080027c3f4a4  782bcb83d2b0  .mntcheck  postinitscripts

/images/dev/0800271fe49c:
.   d1.fixed_size_partitions  d1.minimum.partitions  d1.original.swapuuids  d1p2.img
..  d1.mbr                    d1.original.fstypes    d1p1.img               d1.partitions

/images/dev/080027aff8e3:
.   d1.fixed_size_partitions  d1.minimum.partitions  d1.original.swapuuids  d1p1.img  d1p3.img  d1.partitions
..  d1.mbr                    d1.original.fstypes    d1.original.uuids      d1p2.img  d1p4.img

/images/dev/080027c3f4a4:
.   d1.fixed_size_partitions  d1.minimum.partitions  d1.original.swapuuids  d1p2.img
..  d1.mbr                    d1.original.fstypes    d1p1.img               d1.partitions

/images/dev/782bcb83d2b0:
.  ..  d1.fixed_size_partitions

/images/dev/postinitscripts:
.  ..  fog.postinit

/images/postdownloadscripts:
.  ..  fog.postdownload

/images/Win10uefi:
.   d1.fixed_size_partitions  d1.minimum.partitions  d1.original.swapuuids  d1p1.img  d1p3.img  d1.partitions
..  d1.mbr                    d1.original.fstypes    d1.original.uuids      d1p2.img  d1p4.img

I also made some screenshots on my settings. I also noted that the default storage is linked to eth0, but when I check with ifconfig -a it’s linked to p2p1, seems it’s not finding it because it’s using the wrong network interface? Can’t check now, not at work anymore.

[fogserver@localhost ~]$ ifconfig -a
lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536
        inet 127.0.0.1  netmask 255.0.0.0
        inet6 ::1  prefixlen 128  scopeid 0x10<host>
        loop  txqueuelen 1000  (Local Loopback)
        RX packets 93585  bytes 70082869 (66.8 MiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 93585  bytes 70082869 (66.8 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

p2p1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.2.122  netmask 255.255.255.0  broadcast 192.168.2.255
        inet6 fe80::3ba2:61b4:1954:f66e  prefixlen 64  scopeid 0x20<link>
        ether d4:be:d9:de:5b:ac  txqueuelen 1000  (Ethernet)
        RX packets 86165  bytes 25189476 (24.0 MiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 113615  bytes 35423513 (33.7 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

[Sebastian Roth]

@iekozz Again not correct to the point (upper case L vs lower case l) but that’s not an issue this time. I think we have enough information now.

First I see the .mntcheck files which I thought might be missing. So I was wrong.

Then I see:

/images/Win10uefi:
.   d1.fixed_size_partitions  d1.minimum.partitions  d1.original.swapuuids  d1p1.img  d1p3.img  d1.partitions
..  d1.mbr                    d1.original.fstypes    d1.original.uuids      d1p2.img  d1p4.img

While in the picture the image path is /images/Windows10v1809UEFI. So either rename that folder in the filesystem to match the setting in the web UI or the other way round.

Plus I noticed that you have “Image Manager” set to “Partimage”. I kind of doubt you created this image using the very old Partimage tool (used in FOG years ago). Probably set this to “Partclone GZIP” but that’s just me guessing here.

I also made some screenshots on my settings. I also noted that the default storage is linked to eth0, but when I check with ifconfig -a it’s linked to p2p1, seems it’s not finding it because it’s using the wrong network interface?

Could be but I don’t see all the important information in the picture to be sure. Post the top part of the storage node setting as well so we see the IP is matching the one we see in ifconfig.

[iekozz]

@Sebastian-Roth Ah yes, I really need my glasses

I have several images, so kinda weird to point to one image folder? Or just move them all in one big folder? Will change it from partimage gzip, thanks for the tip. Here is the picture you requested:

[Sebastian Roth]

@iekozz said in Unable to locate image store:

I have several images, so kinda weird to point to one image folder? Or just move them all in one big folder?

In FOG each image needs to be in a separate subfolder residing in /images/<IMAGENAME> plus a matching image definition in the web UI pointing to exactly the right path. In the output you posted I can only see one single image that is complete to be used. The paths did not match and so I suggested to correct that either by renaming the path in filesystem or in the image settings. As well you have a couple of images in subfolders in /images/dev sitting which were not fully uploaded. That’s another story to be told. Let’s try to get that one single image to deploy again and then move on.

[iekozz]

@Sebastian-Roth thanks, got it fixed by renaming the folder. I tried to image it and that worked, but it only boots on drives without data on it.

If data is on the drive, it returns a error with non bootable media found.

[Sebastian Roth]

@iekozz said in Unable to locate image store:

If data is on the drive, it returns a error with non bootable media found.

What do you mean by that? Explain in more detail.

[iekozz]

@Sebastian-Roth Here is what I do.

1: start fog pxe menu
2: deploy image
3: fog starts image process and completes
4: when the pc reboots, it displays a message with no bootable devices found.

This happens if there was another windows install on the disk. Even though it is deleting the tables when imaging.

When i delete everything from the disk, it works, and starts windows.

[Sebastian Roth]

@iekozz said:

When i delete everything from the disk, it works, and starts windows.

Have you tested this on the very same machine? I kind of doubt as we do clean the partition tables as well.

Sounds like an issue where differences in machines play a role. Lets start with the obvious, are both machines legacy BIOS setups or UEFI or mixed?

[iekozz]

@Sebastian-Roth Yes, on the same machine. Also, was using a different pc (Dell optiplex 990) and even when deleting all the data, it fails with a non boot device found error. But it does copy everything to disk.

This is a uefi system using a uefi image.

Also, when I make a new image, the folder is not created under /images, it does save it somewhere, no idea where.

[Sebastian Roth]

@iekozz said in Unable to locate image store:

Also, when I make a new image, the folder is not created under /images, it does save it somewhere, no idea where.

This is a different issue. Images are stored in subfolder in /images/dev/… named like the client MAC address. The folder should be renamed/moved to /images/<IMAGENAME> after it finished uploading. As this does not happen I am fairly sure your FTP connection is not setup properly anymore. The installer does this for you but if you changed the fog user password it will cause that kind of problem. Do you know that user’s password? See if it matches the one you have in the config files /var/www/html/fog/lib/fog/config.class.php and /opt/fog/.fogsettings

[iekozz]

@Sebastian-Roth I see the password I made at the installation, never changed a thing from the default. I see my password in those two files, they are the same, or do you mean the mysql password? I left that blank, easier for me.

Did some further digging:

1: ftp password is set in the tftp settings and my storage settings
2: ftp path is set to /images
3: everything is owned by the fog server
4: the firewall is allowing all traffic
5: the ftp service is running

Only thing I have not tested it a test file to send to the ftp server. The one thing I did notice though is that when the image capture is done, it can’t update the database because the utc time is not set (or something like that) could that be it?

[Sebastian Roth]

@iekozz said in Unable to locate image store:

Only thing I have not tested it a test file to send to the ftp server.

Please do exactly that to see if you really have the right credentials. From what you said I have the feeling that we are talking about different things. The installer creates a web interface user called fog but that is not the same. The FTP user fog is a local Linux account and has a different password set. A longer random password not displayed by the installer.

The one thing I did notice though is that when the image capture is done, it can’t update the database because the utc time is not set (or something like that) could that be it?

Could play a role. See here: https://forums.fogproject.org/topic/12036/centos7-fog-1-5-3-image-capture-fails-with-incorrect-date-timezone-info/2

[iekozz]

@Sebastian-Roth Now that you mention it, there is another user on my fog server, named fog. I’ve never set-up a password for that account though? Where can I find/change it?

[Sebastian Roth]

@iekozz The account fog is create by the FOG installed and should not be touched.

Password reset is done as usually in Linux passwd fog. But if you change the password FTP will cause issues which is used in the background.

[iekozz]

@Sebastian-Roth Ok, so I can just change the password from the terminal and then enter that password everywhere in the ftp settings?

Strange that this is not displayed in the installer or you can choose it somehow.