PDQ Deploy Push/Pull speeds the same
I'm not able to figure out why my Push and Pull speeds are exactly the same: 2MB/sec.
When I use Push, it is from my own PC of course, and the speeds are the same as when it is Pulling to the client from a DFS network share created specifically for this purpose.
I'm not seeing what else I should check. Perhaps a bad brain day?
1
Comments
Todd,
What computer are you test deploying to? You could be limited by the network connection speed of that computer. 16 Mbps could easily be a low-end WiFi connection, for example.
Are the file server, your PC, and the test computer all on the same network/at the same site, or is there traffic traveling over a WAN link?
Setup:
My laptop talks to PDQ Deploy Enterprise on Server. Wired connection.
PDQ Deploy Enterprise Server talks to target computers. Wired connection.
I don't use wireless for my PCs in deployment, nor do I use it in the field, when possible.
This occurs over the WAN and at the same site, too.
Todd,
"Do you have the "Limit bandwidth usage to:" field set in Preferences > Performance?
Using Pull mode does not let me set that.
Using Push mode gives me the option, but (if I understand this right) I set it to 100% so there is no functional limit.
Todd,
You mentioned that you are using a DFS share for your file repository. How many file servers are serving this DFS share, and are they spread across several sites? Is one of the file servers that is serving this DFS share located on the same local network as the computer that you are test deploying to?
So far (basically until I get this working at a good speed) we're only using the DFS share at one site, and it is the same local network as the test deployment.
Todd,
Ok, I think that's probably why you're getting the same speed for both, then. Per PDQ's help pages on Push vs Pull deployments, the benefit of using "Pull" mode is when you are copying files over a WAN link. See this snippet from the help sheet:
They also have a good Youtube video about the two modes and the pros/cons of each: https://www.youtube.com/watch?v=PzwHhkDqZGg
You could also try testing your deployment to several computers at the same time on each mode. If you do that, I would expect Pull to perform better than Push. If you are deploying to only one machine, you probably won't see a performance difference.
Hmm. Okay. That makes sense.
Would you happen to have advice on how to find out why I'm only getting a 2MB speed either way, or should I go buy my network guy a beer and ask him? :)
That's hard to say without knowing more about your environment. I would try installing wireshark and iperf on the file server and the test deployment computer to do some testing. Run a speed test between them using iperf to make sure the network link between them is capable of the higher speeds:
https://iperf.fr/iperf-download.php#windows
Then start a packet capture on both machines before you run the deployment, stopping the packet capture after the deployment is done.
I'm assuming that the test deployment computer and the file server are on different subnets? Is there a firewall between them?
Different subnets and a firewall, yes. I'll do some packet work and see if I can find it!
It's possible the firewall is doing something to decrease that speed. QoS or something like that.
You could try spinning up a test PDQ server and copying your test package to it on the same subnet as the test deployment computer and then pushing it from there-- since they'd be on the same subnet it would bypass any layer 3 stuff going on such as a firewall.