FOG continuously syncs files that haven't changed to remote storage nodes

[kafluke]

I’m using windows based machines for the storage nodes using NFS server. FOG is only running on the master node and the HDD is fine.

[Tom Elliott]

@kafluke so this is where things get interesting. Storage nodes use a common web file that tells them which files have which sizes and hashes. if the nodes don’t have the webfiles this would explain the continuous syncing. Fog can’t validate the remote files are the same so it will continuously delete and resync all the files.

[george1421]

@tom-elliott said in FOG continuously syncs files that haven't changed to remote storage nodes:

. if the nodes don’t have the webfiles this would explain the continuous syncing. Fog can’t validate the remote files are the same so it will continuously delete and resync all the files.

Ugh, this is a game changer. When did this “feature” change? (not intending to make this sound harsh) This requirement will now break using a nas as a storage node.

[kafluke]

@tom-elliott Is there any way to get these webfiles transferred over to a windows NAS? I was able to move the hidden linux files that are found in the image directory. When FOG syncs where does it look for these webfiles? I can mount them in a NFS directory if need be.

[george1421]

@kafluke Assuming its possible, you will need to install php on your windows nas. I started working on a PoC for using a windows 2012 server as a nas. You can either install WAMP (well just WAP) on windows or install a php plugin for IIS.

[kafluke]

@george1421 Setting up a webserver on 40+ sites is not really a viable solution. I may have to pitch some cheap boxes for nodes only and send them to the sites. Thanks anyway guys. This may be worth noting on this link

https://wiki.fogproject.org/wiki/index.php?title=Windows_Storage_Node

that replication will continuously occur over and over again.

[Tom Elliott]

@george1421 it was changed because bashing files was using a lot of bandwidth when having The main do the work. Now we make a request to the node, the node then passes back the hash of the file it’s looking for, leaving only minimal bandwidth usage.

[george1421]

@tom-elliott I think that the new method is cleaner than before. Is that call in the fog replicator code? The thought would be to reverse engineer it. Or see if we could get some node.js code running on the NAS to respond to the replicator with the answer its looking for. I know quite a few people use a NAS as a storage node.

[Tom Elliott]

The code in php is:

     /**
      * Returns hash of passed file.
      *
      * @param string $file The file to get hash of.
      *
      * @return string
      */
     public static function getHash($file)
     {
         usleep(50000);
         $file = escapeshellarg($file);
         $filesize = self::getFilesize($file);
         if ($filesize <= 10485760) {
             return trim(
                 shell_exec("sha512sum $file | awk '{print $1}'")
             );
         }
         return trim(
             shell_exec(
                 sprintf(
                     "(%s -c %d %s; %s -c %d %s) | sha512sum | awk '{print $1}'",
                     'head',
                     10486760,
                     $file,
                     'tail',
                     10486760,
                     $file
                 )
             )
         );
     }

The function is not necessary to be in a class file. The url that gets called is: http://{nodeip}/fog/status/gethash.php

The passed data (the $file parameter) comes from the sent link: ($_POST['file']).

If you can create a file on the storage node in /path/to/web/files

mkdir -p fog/status

Create a php file named gethash.php

Add the lines:

<?php
// Delay for 50 milliseconds.
usleep(50000);
// File is passed base64 encoded, decode it here.
$debase_file = base64_decode($_POST['file']);
// We can't trust user inputs, so escape the file we will check.
$file = escapeshellarg($debase_file);
// The file doesn't exist, return blank.
if (!file_exists($file)) {
    return '';
}
// Get the filesize.
$filesize = trim(
    shell_exec("du -b $file | awk '{print $1}'")
);
// If the file is less than one meg, return the hash of the full file.
if ($filesize <= 10485760) {
    return trim(
        shell_exec("sha512sum $file | awk '{print $1}'")
    );
}
// Otherwise return the first and last meg hashed -- hashing full files takes a long time
return trim(
    shell_exec(
        sprintf(
            "(%s -c %d %s; %s -c %d %s) | sha512sum | awk '{print $1}'",
            'head',
            10486760,
            $file,
            'tail',
            10486760,
            $file
        )
    )
);

This of course assumes a few things.

FIrst:
PHP is installed on the NAS (probably is)
du is installed and available.
awk is installed and available.
sha512sum is installed and available.
head is installed and available.
tail is installed and available.

Hopefully this helps.

[george1421]

@tom-elliott Now that’s service!!

I can say for the synology nas, you have to install php as a package. I’ll confirm that the other utilities exist.

[george1421]

@george1421 Confirmed these applications exist on my test synology nas (I’ve done bad things to this NAS so it might have the kitchen sink too).

du is installed and available.
awk is installed and available.
sha512sum is installed and available.
head is installed and available.
tail is installed and available.

They also exist on my production synology nas.

[kafluke]

Went back to Fog Node’s instead of Windows and it’s working as expected now. You can resolve this.