Continuous Fog Storage Node Replication problem

[MPSAdmin]

@wayne-workman Apologies for not providing clearer info. The paths for the storage node are configured within fog as follows:

The ftp paths are as follows:

The files being replicated are within the G3-0118.1 folder, as listed below:

the source folder on the fog server is the /images folder, as shown below.

Additionally, I have verified that both the synololgy and the fog server are using the same ntp source as the rest of the devices on the network. (a windows domain controller). After changing the ntp settings, i stopped the fog replication service, deleted the files from the synology folder, and restarted the replication service.

Thanks. I’ll keep digging. Most likely it is something simple that I have done.

[MPSAdmin]

@george1421 yes, the files are making it to the NAS. Fog is replicating at full speed, up to the configured bandwidth limit. Once the files are replicated, the process begins anew.

[MPSAdmin]

Also, just checked with the d1.mbr file… md5 checksums match.

040d5ad57af1942104ab788d3b12778e fogserver
040d5ad57af1942104ab788d3b12778e synology web console

d1p1.img…checksums match here too.
79d8b1a77b8111418481c121c246501b fog server
79d8b1a77b8111418481c121c246501b synology web console

[george1421]

@mpsadmin OK then we might need an opinion from one of the @developers to give us an idea why the NAS is not reporting what the FOG replicator expects.

If I remember right there is 2 or 3 replicator logs. One is the main replicator log, one is for the storage node, and there may be a log file for each image being replicated. I don’t have access to my production FOG server at the moment to confirm. But I thought the logs were pretty descriptive to why it was replicating.

[george1421]

@mpsadmin I have my fog-pi server with me that has FOG 1.4.4 on it. I’m looking at the code and it appears to rely on the file byte count being the same to indicate if the file needs to be replicated or not. I’m not a programmer so this object oriented php is a bit greek to me. But so far it looks like byte count is the trigger. Now this is the byte count as reported by FTP session. Also the log writes to /opt/fog/log look fairly descriptive to what exactly the replicator is doing.

[Wayne Workman]

@mpsadmin Did you use the FTP option when logging in via WinSCP? I ask because I want to be sure. Photo:

[Wayne Workman]

@mpsadmin said in Continuous Fog Storage Node Replication problem:

the portions I deleted from the originall fogreplicator.log are html pages from the synology. Is there a reason that the fog server is pulling html files from the synology?

@Developers I believe that quote is a symptom of the root problem. I am just trying to drill down to why.

[Tom Elliott]

@wayne-workman It’s not a problem fog is causing. It’s a problem with how fog manages to do things. The request being made to the synology is not a fog based system and so synology is responding with its own information.

This is just a guess on my part, but it’s unlikely to be anything we can fix.

[MPSAdmin]

@wayne-workman yes. used ftp.

[MPSAdmin]

Here are the logs from the replication, if you have time to look at them.

fogreplicator.log https://pastebin.com/Lb0Kc17n
fogimagesize.log https://pastebin.com/NN0nyrbY
fogreplicator…transfer.spg-syn1.log https://pastebin.com/VHtHRZq3
fogreplicator.G3-0118.1.transfer.spg-syn1.log https://pastebin.com/U51QxtkY

[MPSAdmin]

What protocol does fog use to verify its images? somehow fog is pulling html from the synology… makes me wonder if it is looking in the /fog web path on the synology?

(the synology guide by george says this folder doesnt matter because its not a real storage node)

wondering if I can ‘create’ a webserver on the synology with the path fog is looking for? just a little more patience guys…i’d appreciate it.

Thanks!

[george1421]

@mpsadmin My intent was to set this back up in my lab but day to day work has got in the way.

From looking at the code it appears that the fog server replication connects to the storage node over FTP and pulls the target file size. If they are not the same then the replicator starts over.

[Wayne Workman]

@george1421 said in Continuous Fog Storage Node Replication problem:

it appears that the fog server replication connects to the storage node over FTP and pulls the target file size.

I don’t think that’s how it’s doing it. I think it’s making an HTTP call to a PHP script on the storage nodes to get file size. And that’s why it’s outputting an entire webpage instead of a size (because the NAS is redirecting the call to it’s landing page). The FOG Image Replicator is supposed to revert to FTP for filesize if the HTTP call doesn’t work. I believe it’s only checking the HTTP status code and not verifying that the output is a single number - so that’s why an entire web page is being pushed to the replication logs - and why it’s not reverting to FTP for filesize. @developers

[MPSAdmin]

so… as an update…
I stopped/disabled the fog replication service and used the same command as fog used to replicate the folder over from the fog images store to the synology images store.

After this, imaging from the storage node works as desired.

I’d really like to know how fog winds up pulling the Synology home page during the file compare… is it using http? what path is it trying to hit? I can set up a packet trace and try to find out, but I’m hoping someone can just chime in with the info.