Web Hosting Talk : Web Hosting Main Forums : Hosting Security and Technology : Rsync problems recently started

[santrix]

Hi... Recently upgraded my vps (this is the only think I can think of), but that means it now has more RAM, more CPU capacity, and more disk space, so I doubt that has anything to do with the new problem.

I am trying to rsync my WHM user account backup directory to a machine in the office.. It's alway run fine in the past (it's something I do every couple of months as an extra extra backup!)

On my mac, I usually run

rsync -avze 'ssh -p 999' root@myhost.co.uk:/drive2/localbackups/ /Mystuff

My Mac runs 2.6.9 so I installed the latest release just to be sure 3.0.6 and put it in usr/local/bin

/usr/local/bin/rsync -avze 'ssh -p 999' root@myhost.co.uk:/drive2/localbackups/ /Mystuff

Each time i try this it gets a few dozen files into the transfer and just dies, and eventually times out with the following...


I watch the server with "top" while I open the connection, and I see rsync fire up and consume reasonable resources, but once it hangs, it just drops to idle.

My WHM server is pretty stock, and is running version 2.6.8

Is it safe to upgrade rsync on the WHM server, or will this cause probs with other scripts that WHM runs?

[srvadmin]

Try to use rsync with option --timeout=0 which means rsync should never timeout.

[santrix]

The error I am getting is

cpbackup/daily/pathtosomefile/file.jpg
Read from remote host vps.myhost.co.uk: Operation timed out
rsync: connection unexpectedly closed (10858655 bytes received so far) [receiver]
rsync error: error in rsync protocol data stream (code 12) at io.c(600) [receiver=3.0.6]
rsync: connection unexpectedly closed (2487542 bytes received so far) [generator]
rsync error: unexplained error (code 255) at io.c(600) [generator=3.0.6]


I'll try the timout thingie, but I don't see why it should timeout in the first place... no load, small files, etc

[santrix]

Nope... still getting

Read from remote host vps.myhost.co.uk: Operation timed out
rsync: connection unexpectedly closed (3885965 bytes received so far) [receiver]
rsync error: error in rsync protocol data stream (code 12) at io.c(600) [receiver=3.0.6]
rsync: connection unexpectedly closed (2487542 bytes received so far) [generator]
rsync error: unexplained error (code 255) at io.c(600) [generator=3.0.6]

[madaboutlinux]

"Operation timed out" suggests more of a firewall issue than anything else. It looks like the remote server is blocking the source for some reason and the connection just times out after a specific time. Try to disable the firewall on both the servers first and if rsync works, then check which rules are causing problems.

[Steven]

It's not a firewall issue.

Are both servers using the same version of rsync?

[santrix]

No it's not a firewall issue...

Well, the client (one of the macs in my office) is running 2.6.something and also 3.0.6 has been installed to try and fix this, but still no luck.

The servers are all running the stock centos 5.x rsync 2.6.8

I am considering upgrading rsync on one of the servers, but am not sure if this is going to upset WHM...

I'm embarrassed to admit that although I use rsync regularly, I'm not entirely sure how the transfers take place... i.e. When I initiate rsync from my mac and connect to a remote server, does rsync have to be running on the remote server in daemon mode in order to receive a connection??? I'm not at all clear on this point.

[rwxguru]

upgrade rsync and retry as for running rsync as service on server side this is not the case since you're using ssh transfer as I see from your command .

[plumsauce]

Originally Posted by santrix

Nope... still getting

Read from remote host vps.myhost.co.uk: Operation timed out
rsync: connection unexpectedly closed (3885965 bytes received so far) [receiver]
rsync error: error in rsync protocol data stream (code 12) at io.c(600) [receiver=3.0.6]
rsync: connection unexpectedly closed (2487542 bytes received so far) [generator]
rsync error: unexplained error (code 255) at io.c(600) [generator=3.0.6]

Without asking who your isp is, try the following search on a search engine of your choice:

isp sandvine

It could be that your isp is using sandvine, and sandvine is chopping the connections.

[Steven]

Originally Posted by santrix

No it's not a firewall issue...

Well, the client (one of the macs in my office) is running 2.6.something and also 3.0.6 has been installed to try and fix this, but still no luck.

The servers are all running the stock centos 5.x rsync 2.6.8

I am considering upgrading rsync on one of the servers, but am not sure if this is going to upset WHM...

I'm embarrassed to admit that although I use rsync regularly, I'm not entirely sure how the transfers take place... i.e. When I initiate rsync from my mac and connect to a remote server, does rsync have to be running on the remote server in daemon mode in order to receive a connection??? I'm not at all clear on this point.

Rsync often goes over ssh

[santrix]

Thanks for the Sandvine heads up - I only recently added a talktalk connection to the office as a backup (which I am currently using), and I forgot to check if there was any traffic management issues. I bet that's what it is... I'll chase up TalkTalk.

On another note, and I know this is a little OT, but I have had a read, and think I get now how rsync communicates....

Without any host in the remote part of the call, the transfer is purely local.

With a host and a colon in the remote part, then we try to open a remote instance of rsync either directly or via ssh, to facilitate the transfer.

With a double colon in the remote call we are explicitly expecting to connect directly to an existing daemon process on the remote side.

Does that sound about right?