04-02-2018 11:53 AM
I have a Windows 10 VM running under ESXi to a USW8>USG. This VM is setup to do off-site sync of my fileserver to cloud storage and thus I have a need to limit it's upload bandwidth. Unifi sees this VM like any other client on the network and is statically assinging an IP to it however at this moment Unifi sees 0bytes of traffic flowing from this VM. My initial solution to capping it's speed was to create a new user group with a 15Mb upload limit and then I assigned the client to this group, however since it doesn't see any of the traffic, it has no way to limit it's bandwidth use.
I can can see, historically, that UNifi has seen some traffic from this client, could it be that the sync to Google is SSL and that traffic in this format is entirely invisible to Unifi? I really do not know.
Thank You in advance.
04-02-2018 12:18 PM
04-02-2018 12:50 PM - edited 04-02-2018 12:52 PM
Attached...you can see it has recorded some data, randomly it will see data flowing, but often it says 0. Also, no, the setup is cable modem>USG>US-8-60W> Clients and a UAP-HD. running 5.7.20 and the firmwares are all the reccomended firmwares for this controller version.
If the data is being uploaded with SSL, is that any issue for Unifi to see that packets are flowing and how big?
04-02-2018 01:30 PM
Is your host a Dell? I get all sorts of weird disconnect/reconnect messages under events, although nothing actually does disconnect, and the host and VMs show 0 bytes for the time it thinks they were connected. I've never had a single thing in Unifi be consistent in the month I've had this stuff, I just ignore it now, and the stats, overview and reporting were a big draw for me.
04-02-2018 03:24 PM
@Zak95 The reason you're seeing disconnect/reconnect for VMs in events is that those VMs have no traffic flowing across any UniFi devices, so they don't get informed to the controller. After 5 minutes the MAC address of the VM drops from the switch's MAC table. When the VM answers a broadcast (such as an ARP request) then the switch will store it again, hence the disconnect/reconnect messages.
Best way to stop that is either a network monitoring system, even something basic like smokeping, or you could run a bash script on the VMs themselves (that's what I do).
04-02-2018 03:53 PM
Ahh thanks I wasn't aware of that I'll look into it, it's not a big issue just annoying. My host shows disconnects too, but it doesn't disconnect, someone else had an issue with a dell ESXi host, thought there was a small can of worms.
2 weeks ago
Sorry to bump this but I'm having the same problem.
Since I've been started using Ubiquity devices, the controller won't recognize any live activity of my (whitebox) ESXi server and its VMs. The controller itself is also located on that ESXi.
The controller does indeed recognize the VMs correctly and tells me the IP and MAC addresses but activity always stays at 0bps. I can see some (correct) historical data when looking at the DPI tab.
My setup is quite simple:
ESXi's VM NIC -> USW8 -> USW8 -> USG3P -> ISP-Router
The switches are running FW 4.0.22, USG is on 4.4.36 and the controller is 5.10.12 ... but it didn't work with stable releases as well.
a week ago