12-02-2019 03:46 AM - edited 07-05-2021 11:22 AM
Hi guys,
I am having some weird problems with my 2504 controller running on 8.5.151.0
I noticed today that I am unable to achieve more than 30-32Mbps on an individual device. It's like there is a hard-cap that doesn't let the device getting more bandwidth.
I have played alot, trying to tweak different QoS settings, but apparently 2500 is very limited in configuration of the bandwidth rates, burts and traffic profiles. I was unable to find a way to completely disable the qos for an SSID and just let it be on "best effort".
My questions are:
* are there any built-in, default policers for different classes and if so - what are the values?
* Is there anyway possible I can completely switch off that behaviour (preferably on a per SSID basis?)
* Is that an 8.5.151.x problem because I don't recall having this problem on older versions
01-14-2020 03:36 PM
01-14-2020 11:40 PM
01-21-2020 01:46 AM
Hey there,
Go to Wireless- Global Configuration and set the TCP MSS to 1250 is one suggestion from my side. Might be having issues with fragmentation which can start increasing the latency. Although strange.
Also get two devices on the wired network in the same VLAN, as the one that you have issues with and do an IPerf test. Make sure to verify that there isn't an issue with the DS.
Scott also suggested the use of Flexconnect local switching. If the WLC has issues this will bypass it for sure. (It will be a good idea to give it a test).
Long ago there was a bug related with those 1850 APs that they were having low throughput, not sure if you are not hitting the same thing.
01-21-2020 02:49 AM
Hi mate and thanks for your response.
I don't think the MSS is playing any significant role here as it is within the local LAN where no overhead is added to the IP header. Furthermore, the tests observed were conducted for both - TCP and UDP and the results are not much better (the only difference kicks in when latency increases during the bandwidth throughput and as a result due to the nature of TCP (being hit by RTT))
Wired devices within the same VLAN perform as expected at line rate (~120MBs/~1Gbps).
I think I have established another problem where my both MACs perform differently, which may come down to a driver issue, but other PCs on the network also cannot achieve the bandwidth I was able to get in the past.
I started reading this guide and will try to follow it as close as possible & will share the results back.
Thanks
01-21-2020 06:25 AM
Boy, you got your hands full there.
Check this best practice guide. Might have some useful information inside.
Couple of other stuff coming to my mind are:
- Check what is the Current TxRateSet for the user when you perform the tests. See if the clients are just not able to jump on higher data rate (that would mean that there are retransmissions).
- See if you have Mu-Mimo enabled on the SSID (if yes - disable it, if no - enable it). There was a bug that was supposed to be fixed on your version with clients that have Mimo in power saving mode - limiting their throughput on 802.11ac. But hey, doesn't hurt to try it.
- Test on 2.4Ghz? See what rates and speeds can you get if you are on 802.11b/g/n.
- Verify the "Interoperability with Other Clients" section in the Release Notes for 8.5.151.0 version. There were macs there with operation version that are expected to have issues with this firmware (which would mean that a client update is required).
Btw, nice guide with the MACs :) I would definitely have it in mind for future references thanks!
Looking forward to hear from you if this was helpful.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide