Community discussions

 
Nefraim
just joined
Posts: 8
Joined: Fri Apr 13, 2018 10:01 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G (not yet for 2.4GHz) and 5G WiFi

Sat Apr 28, 2018 11:59 am

How about yours in these figures (ping time and transfer rate)?
2.4 GHz band is very dirty. I can not do a clean experiment. My hap ac^2 is receiving a signal from 46 stations near. Best results on channel 13, average 49 Mbit/sec. On channel 6 average 24 Mbit/sec and very unstable. Therefore, 40 MHz can not be used with hi speed. Maybe a mikrotik can improve noise immunity. But such tests as at 5 GHz, at 2.4 GHz in the house can not be carried out.
If you experience different speed when changing 2.4 GHz channels from 1 to 6 or 13 is pointing to another emitter in your area. Mikrotik can try to improve transfer rates but usually in crowded scenarios, changing the channel will be the best solution afaik. This is why you get better result when you changed the channel.

You might want to try to see if another electric equipment isn't messing the 2.4 Ghz channel. My worst experience was when a stupid 2.4 GHz wireless mouse was interfering with my signal. I was blaming the router turns out it was the stupid mouse wireless receiver that was hampering may connection. If I had my wireless mouse on my connection was garbage unplugging the thing everything was fine. Try to see if you don't have any other emiters neear by.
 
Traveller
just joined
Posts: 24
Joined: Thu Apr 05, 2018 10:12 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G (not yet for 2.4GHz) and 5G WiFi

Sun Apr 29, 2018 11:12 am

If you experience different speed when changing 2.4 GHz channels from 1 to 6 or 13 is pointing to another emitter in your area. Mikrotik can try to improve transfer rates but usually in crowded scenarios, changing the channel will be the best solution afaik. This is why you get better result when you changed the channel.
Yes. But, for 40 MHz bandwidth i must use two channels with a distance 4 channel: 1 and 5, 6 and 2, 6 and 10 or 9 and 13. I do not have two free channels. I have free only 13 channel. On free channel hAP ac^2 transmits data without failures.
You do not have the required permissions to view the files attached to this post.
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon Apr 30, 2018 6:21 am

After some testing this weekend with a Realtek USB AC adapter and a single client device, I found the following trying to get the best I could from the router without heavy tweaking of the Wireless settings:

  • 2.4 GHz: no significant performance difference from RouterOS 6.42.1 to 6.43rc5. Approx. 130 Mbit/sec download, 85 Mbit/sec upload on average.
  • 5 GHz: up to a 15% performance improvement from RouterOS 6.42.1 to 6.43rc5 (includes running the routerboard firmware upgrade). Approx. 400 Mbit/sec download, 350 Mbit/sec upload.
  • very low 10 Mbit/s UDP stream losses (<0.5%), low to very low jitter (<10ms), zero standard 1s ping packet losses, few large pings (<2%)

Test conditions:

  • low interference, according to spectral scans
  • 1m distant from the wireless client to test router
  • all tests were through the router interfaces (WAN-WLAN) using PC endpoints
  • regulatory domain TX power for 2.4 GHz, default manual-tx-power for 5 GHz

More details on the 6.43rc5 2.4 GHz band status:

  • TX/RX Signal Strength: -25 to -35 dBm
  • Signal to Noise: 70 to 80 db
  • Tx/Rx CCQ: 85% to 100%
  • Tx Rate: 300 Mbps @ 40 MHz
  • Rx Rate: jumps around a lot, anywhere from 6 Mbps to 216 Mbps@40 MHz

So, overall it appears the Mikrotik hAP ac^2 is very good at 5 GHz now for a 2x2 device, especially its upload throughput.

The 2.4 GHz performance could be more symmetrical. That upload could perhaps be improved by another 15%, if the Netgear Orbi is anything to go by.

I'm still intrigued by the 2.4 GHz latency and stability. There seem a lot of conflicting reports. But from my short (<10 minute) tests, I could not find any significant flaws.

I guess there could also be compatibility issues remaining with certain client WiFi chips, given what's happened in the last two patches.
 
User avatar
DanielJB
Frequent Visitor
Frequent Visitor
Posts: 53
Joined: Mon May 27, 2013 3:05 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon Apr 30, 2018 11:53 am

1. where there is spectral congestion (eg 2.4GHz networks), I suggest using hw-protection-mode=cts-to-self and hw-protection-threshold=200 (to keep overhead low on TCP ack packets)
2. hw-protection-mode=rts-cts has higher overhead but may give better client->AP throughput if the AP transmit power is significantly higher than the client (100mW vs 30mW) as nearby APs will be able to hear the AP's 'client about to transmit' message
3. conduct spectral scanning to identify low-energy/interference bands (5GHz possible on older Mikrotik 802.11n 5GHz radios)
4. you will see better throughput if SSIDs overlap fully (eg on 2.4GHz, all are on channel 1, 6 and 11) as the collision probability would be lower
5. where there is significant congestion, reduce channel width to minimise retransmissions due to collisions. I feel Mikrotik should make channel width 20MHz default @ 2.4GHz, no?
 
alengrosevic
just joined
Posts: 10
Joined: Fri Mar 30, 2018 12:14 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon Apr 30, 2018 2:11 pm

@squeeze: Colud you please explain to me how did you, based on your test on distance of 1m, concluded that wireless part is good? On distance of 1m, I would use cable, this is not the type of test that is used for wireless. Put distance of 5m, then 10m, and even if you can 15m between devices. Then put on wall between them. Then you will get real scenarios. Distance of 1m is maybe to see max throughtput od wireless, but if you want to see if this device is good for home/office use, then please test it diferently. In my case, RC5 did improve 5GHz performance, but this is still not finished. After using several AC devices before HAP AC2, I can tell that performance of Wireless part is still behind other competitors.
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon Apr 30, 2018 2:30 pm

@squeeze: Colud you please explain to me how did you, based on your test on distance of 1m, concluded that wireless part is good? On distance of 1m, I would use cable, this is not the type of test that is used for wireless. Put distance of 5m, then 10m, and even if you can 15m between devices. Then put on wall between them. Then you will get real scenarios. Distance of 1m is maybe to see max throughtput od wireless, but if you want to see if this device is good for home/office use, then please test it diferently. In my case, RC5 did improve 5GHz performance, but this is still not finished. After using several AC devices before HAP AC2, I can tell that performance of Wireless part is still behind other competitors.

I did these basic tests because even such had awful performance in multiple aspects before. Note the thread title and past posts ...

Nor was I interested in doing any broader tests of the wireless of this device until such basics worked. There was little point wasting time on it if even the simplest context I could create produced no improvement. Now since it appears worth it, I can test further at my leisure with my use cases, which would be more difficult to compare with others anyway.

If you would like to contribute with your own "real scenarios" tests or explain what you mean by "performance of Wireless part is still behind other competitors", feel free.
 
User avatar
Davis
Frequent Visitor
Frequent Visitor
Posts: 64
Joined: Mon Aug 01, 2011 12:27 pm
Location: Latvia, Riga
Contact:

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Tue May 01, 2018 2:32 am

I have noticed that (at least in my specific setup) hAP ac² together with Intel 7260 wifi card has issues with action=set-priority mangle rules.
I had used this mangle rule (rule comes from this MUM presentation) and disabling it in some cases drastically improved performance (e.g. SFTP file download improved from 1-2 MB/s to 9-10 MB/s limited by internet speed):
/ip firewall mangle
add action=set-priority chain=postrouting comment="DSCP to 802.11 WMM" \
    new-priority=from-dscp out-interface=bridge passthrough=yes
Also I have noticed that different versions of Intel drivers behave differently (e.g. 16.5.3.6 seem to offer better and more stable speeds than 18.33.12.2).
 
pepek
just joined
Posts: 23
Joined: Tue Apr 10, 2018 12:14 am
Location: CZ

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Tue May 01, 2018 8:10 am

To anyone (yhfung, strods?) who marked this topic as SOLVED - could you un-mark it please? This problem is not so bad as before, but it is not solved.

Today, I made another test and it showed same speeds as last time (even slightly slower) - but this time, I've disconnected all other wireless devices in office (Access points, phones, wireless mouse...). Also updated Intel WiFi driver to latest one (20.40.0.4 from February 2018 for Dual Band AC 8265). 5G is all clear, 2G gets few networks with RSSI -80dBm, my is -15dbm (in inSSIDer, registration list shows -30dBm for my client)

During this test, I've been disconnected from hAP ac^2 few times, ping is unstable and 802.11ac download is barely faster than 802.11n upload - this is not OK.
2G download speed is half of 2G upload.

SMB server I used is able to feed 1gbps ethernet.
AP setup - all interfaces in bridge, DHCP client on Bridge, band set to only-N (40mhz Ce) and only-AC (80mhz Ceee), protocol 802.11, WPS off, country Czech Republic, WPA2-PSK AES

Ping without any wireless traffic:
Reply from 192.168.7.7: bytes=32 time=12ms TTL=128
Reply from 192.168.7.7: bytes=32 time=19ms TTL=128
Reply from 192.168.7.7: bytes=32 time=16ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=21ms TTL=128
Reply from 192.168.7.7: bytes=32 time=24ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=283ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=2ms TTL=128
Reply from 192.168.7.7: bytes=32 time=3ms TTL=128
Reply from 192.168.7.7: bytes=32 time=18ms TTL=128
Reply from 192.168.7.7: bytes=32 time=18ms TTL=128
Going to send this info to support mail.
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Tue May 01, 2018 8:28 am

This was not marked by me. I am not sure who marked it resolved.

By the way, have you updated your RouterOS and firmware of hAP ac^2 to at least v6.43rc5?

Now I am able to connect my remote AP station formed by an hAP ac^2 (R1) by another hAP ac^2 (R2) with two layers of concrete walls in between and both R1 and R2 are at a distance of around 10 meters. R2 was connected to AP of R1 using bridge mode. I used my MacBook Pro Retina to connect ether3 of R2. The download speed that I can achieve 168Mbps.
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-300.01 sec  5.86 GBytes   168 Mbits/sec  648             sender
 
pepek
just joined
Posts: 23
Joined: Tue Apr 10, 2018 12:14 am
Location: CZ

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Tue May 01, 2018 8:44 am

Yes, I have updated both - ROS and firmware to 6.43rc5

Can you rename the topic? Add "[NOT-" to the end of topic name? :)

I believe Mikrotik has their test setup with good speeds, but they have serious problems with some clients. I suspect all Intel cards. I'm able to reach 100mbps on single chain 802.11n Airlive AP with Intel client, but i have problems reaching this speed on dual-band Mikrotik. This is something Mikrotik has to work on. Probably cooperate with Intel?

Your test is 2G or 5G?
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 9:03 am

Already done as per your request. They were connected via 5G ac link. If you find something strange, besides putting information here, you can also sent it to the Support for help.

YH
 
pepek
just joined
Posts: 23
Joined: Tue Apr 10, 2018 12:14 am
Location: CZ

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 9:07 am

Mail already sent to support, received automatic reply and waiting...
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 10:07 am

It is very strange that I got very good results with my CHUWI notebook, in which Intel ac chip is used. However I forgot the model number. The download speed can reach over 200Mbps. Now I can get only 20Mbps. I have to find out the method to replicate the problem such that the Support Team is able to locate where gives the problem. My MacBook Pro Retina could give me 380Mbps without any difficulties.

YH
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 12:31 pm

Since I have four hAP ac^2, I can very whether it could happen only one of them or all of them. Based on the results obtained below, it shows clearly that the issue of download issue has not been resolved. The maximum download rate (File server (iperf3 server) -> WAN -> WLAN -> Laptop PC (MacBook Pro Retina 2015, CHUWI notebook 2017).

As seen in Listing 1, the download speed is only 25Mbps, which is much below 380Mbps of my MacBook Pro Retina 2015. MikroTik Team has to adjust system code to cater for Intel-basd AC adapters.

YH

Listing 1: Download Test on 5G hAP ac^2 with v6.43rc5 RouterOS uisng CHUWI ac interface (Intel Dual-Band Wireless-AC 3165)
Accepted connection from 192.168.55.172, port 51338
[  5] local 192.168.55.203 port 5201 connected to 192.168.55.172 port 51339
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.01   sec  3.25 MBytes  27.1 Mbits/sec    0    170 KBytes
[  5]   1.01-2.01   sec  3.00 MBytes  25.2 Mbits/sec    0    268 KBytes
[  5]   2.01-3.00   sec  3.06 MBytes  25.8 Mbits/sec    0    271 KBytes
[  5]   3.00-4.01   sec  2.76 MBytes  23.0 Mbits/sec    0    271 KBytes
[  5]   4.01-5.01   sec  2.76 MBytes  23.1 Mbits/sec    0    271 KBytes
[  5]   5.01-6.01   sec  2.76 MBytes  23.1 Mbits/sec    0    271 KBytes
[  5]   6.01-7.00   sec  3.06 MBytes  25.8 Mbits/sec    0    271 KBytes
[  5]   7.00-8.01   sec  2.76 MBytes  23.1 Mbits/sec    0    271 KBytes
[  5]   8.01-9.01   sec  2.76 MBytes  23.1 Mbits/sec    0    271 KBytes
[  5]   9.01-10.01  sec  2.76 MBytes  23.1 Mbits/sec    0    271 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.02  sec  28.9 MBytes  24.2 Mbits/sec    0             sender
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 2:27 pm

Listing 1: Download Test on 5G hAP ac^2 with v6.43rc5 RouterOS uisng CHUWI ac interface (Intel Dual-Band Wireless-AC 3165)
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.02  sec  28.9 MBytes  24.2 Mbits/sec    0             sender

You get no "Retr" from your TCP reverse tests? Or are you not using the "-R" iperf3 option?

Maybe you have an MTU problem. Try the following in a command or Powershell:

ping myserver –f –l 1472
ping myserver –f –l 1470
ping myserver –f –l 1460
etc.

Keep reducing the buffer size until you find one that does not bring up an error. Then set the highest you get on that specific client device with:

netsh interface ipv4 set subinterface “wi-fi” mtu=xxxx store=persistent

where xxxx = number you found.
Last edited by squeeze on Tue May 01, 2018 5:02 pm, edited 3 times in total.
 
User avatar
vecernik87
Long time Member
Long time Member
Posts: 642
Joined: Fri Nov 10, 2017 8:19 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 2:38 pm

My speed grealty improved after enabling "throughput booster" in Intel's wireless settings:
device manager -> network adapters -> [your wireless card] -> advanced -> troughput booster -> enabled
I am not certain what does it do but it may have something to do with timing and colission handling. It is said, that enabling will improve your speed if equal speed for all clients is not required.
Maybe Intel simply waits too much in order to let everyone around communicate equally?
 
pepek
just joined
Posts: 23
Joined: Tue Apr 10, 2018 12:14 am
Location: CZ

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 2:43 pm

No, this setting was not useful for me. (maybe a little, but not significantly)

I had the same speed problems with my previous notebook - Latitude with Intel 7265. I had to replace wlan card to bcm model to get better speed with mikrotik. There was no other speed problem than with mikrotik AP. And I don't want to start changing internal devices on brand new precision...
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 4:05 pm

My speed grealty improved after enabling "throughput booster" in Intel's wireless settings:
device manager -> network adapters -> [your wireless card] -> advanced -> troughput booster -> enabled
I am not certain what does it do but it may have something to do with timing and colission handling. It is said, that enabling will improve your speed if equal speed for all clients is not required.
Maybe Intel simply waits too much in order to let everyone around communicate equally?
Thank you for your suggestion, I did follow your idea to adjust the settings. It turned out that the speed boosted to 220Mbps. When I disabled the settings, it still gave me 220Mbps download. It is really strange and not able to be explained. If you found it did not work, try to reboot your PC.

YH
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Tue May 01, 2018 4:11 pm

Listing 1: Download Test on 5G hAP ac^2 with v6.43rc5 RouterOS uisng CHUWI ac interface (Intel Dual-Band Wireless-AC 3165)
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.02  sec  28.9 MBytes  24.2 Mbits/sec    0             sender
You get no "Retr" from your TCP reverse tests? Or are you not using the "-R" iperf3 option?
I did use the -R option when carry out the iperf3 test. Since the test only ran for a few seconds, it did not have any packets retry for a very short period of time. In the another post, I mentioned that I could obtain higher throughput (220Mbps on Intel-based AC) by applying the suggestion given by other member.

If I have time, I shall follow your mentioned to adjust the MTU to see whether it has any improvement when the speed drops to 2xMbps.

YH
 
alengrosevic
just joined
Posts: 10
Joined: Fri Mar 30, 2018 12:14 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Wed May 02, 2018 12:36 am

I have managed to get some improvements with AC2. I use AC2 on his stand to be installed vertically on the table (5m away from computer). Antennas (X99A Titanium MSI) of my AC wireless card were until now verticaly mounted. Today i have tried to mount them horizontally, just for testing. Signal quality has jumped instantly from 70-75% to 90-95%. Transfer is faster and more stabile (even). Connection speed between computer and router has arised from 520 mbit to 780 mbit. Maybe this information can help someone with this problem.
 
yottabit
Member Candidate
Member Candidate
Posts: 160
Joined: Thu Feb 21, 2013 5:56 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Wed May 02, 2018 2:29 am

The internal antennas in the hAP ac2 are 90-degrees offset. Try aligning your wireless adapter antennas the same way (1 horizontal, 1 vertical, 0/90 degrees; then +45/-45 degrees), and see if this can make further difference.

Sent from my Pixel XL using Tapatalk


 
morph1
just joined
Posts: 9
Joined: Wed Apr 25, 2018 3:50 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Wed May 02, 2018 11:36 am

i think you should hold your horses. You guys did improve the situation, no doubt. But wouldn't call this issue to be resolved:
photo_2018-05-02_10-35-29.jpg
We are getting somewhere, but as you can see in comparison with other products there is still some room to get even better.
Are we expecting more work on this side or is this the best that can be done?

P.S.: still using the same settings/methods as in my previous post
You do not have the required permissions to view the files attached to this post.
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Wed May 02, 2018 2:40 pm

morph1,

Many thanks for showing your revised chart including the firmware upgrade on hAP ac^2.

Based on your chart figures, it shows clearly that the ping time is much larger than others. Because of the poor performance in wireless section (not only on hAP ac^2 prodcuts but all MikrotiK WiFi products), there is a shift of products from MikroTik to Ubiquiti. As for me, I of course want MikroTik to improve their WiFi products as quick and as soon as possible.

In the beginning of this thread and other threads, only few people found satisfactory with their hAP ac^2 but majority of us found big problems on hAP ac^2. Why? How comes? I do believe that the WiFi testing on hAP ac^2 in MikroTik was not complete thoroughly. Nevertheless, the major issue was over but minor tuning has to be continued as some members still founds problems in their WiFi adapters.

YH
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Wed May 02, 2018 2:58 pm

P.S.: still using the same settings/methods as in my previous post

Thanks for the tests. I don't understand the new bottom scale, compared to your previous ones. Is it still 10ms pings and 60k ping count?
 
morph1
just joined
Posts: 9
Joined: Wed Apr 25, 2018 3:50 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Wed May 02, 2018 6:08 pm

P.S.: still using the same settings/methods as in my previous post

Thanks for the tests. I don't understand the new bottom scale, compared to your previous ones. Is it still 10ms pings and 60k ping count?
uhh my bad. totally overseen that we changed the plots :) still using 10ms interval on the pings, 60k of them. so the scale is in [ms]!
The new plot style includes ALL packets.
Since we found out that there is a huge difference in the distribution of the ping responses, we decided to adapt the scale to have all of them included - instead of just the number >10ms.
Hope this helps.
 
alengrosevic
just joined
Posts: 10
Joined: Fri Mar 30, 2018 12:14 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi [Not]

Fri May 04, 2018 12:10 pm

The internal antennas in the hAP ac2 are 90-degrees offset. Try aligning your wireless adapter antennas the same way (1 horizontal, 1 vertical, 0/90 degrees; then +45/-45 degrees), and see if this can make further difference.

Sent from my Pixel XL using Tapatalk
I have tried all the combination but the best result was with the antennas positioned horizontally.
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Sun May 06, 2018 2:58 am

I decided to adapt morph1's low latency testing approach, but for best case scenario, relative stability tests. These were my results.

TL;DR. Mikrotik hAP ac2 has a stability issue with 5 GHz band at 40 MHz channel width (visible also on morph1's results), but all other widths and bands performed well. The original Mikrotik hAP ac has a surprisingly disappointing 2.4 GHz band stability at 20 MHz channel width (often required for those in high interference environment and still using 2.4 GHz devices). Asus are better than I thought, almost a decade ago. :D

Box-and-Whisker Plot of 56-byte data, 10 ms interval, 60k count ICMP pings with three different devices: Mikrotik hAP ac2, Mikrotik hAP ac and Asus RT-N56U as a basic comparison.

I tested 2.4 GHz (default and 20 MHz), and 5 GHz (default, 40 Mhz and 20 MHz) with hAP ac2's wired LAN-to-WAN performance as reference. Default 5 GHz channel width on Mikrotik devices was 80 MHz.

Lower is better:
Image


Standard Box and Whisker parameters used. I don't think morph1 uses the default parameters or has a very different test environment, so cannot compare directly. Analysis tool is Matlab.

The blue box represents the traffic jitter, i.e. non-outlier traffic variation on the connection. Large box = large jitter which is worse for low latency applications.

The red crosses represent outliers that fall so far outside of the total range of the normal traffic, that they can be statistically rejected as random with a 99.3% confidence when the underlying distribution is or tends to the Normal distribution.

The ping range is carefully chosen. A maximum of 40 ms variation is normally desired for streaming applications like VoIP. A maximum of 250 ms delay for other applications like gaming. I zoomed in to 250 ms instead of 500 ms RTT because the worst offenders look the same. The actual maximums observed were sometimes much higher than that, e.g. 2000 ms (2s) for the "ac2 5GHz @ 20/40MHz" configuration and 525 ms for the "ac 2.4GHz @ 20MHz" configuration.

I could have zoomed in to show the jitter more closely for low latency differences, but given the number and distribution of outliers, that is far less important for stability.

Due to the significantly poorer performance of the "ac2 5GHz @ 20/40MHz" and "ac 2.4GHz @ 20MHz" configurations, I actually did them twice to check and took the best run of each. There was no significant difference in either standard deviation or outlier count between these alternate runs.

Test setup is the same as before:
  • PC (client) -> WLAN -> NAT -> WAN (single hop) -> PC (server). No load on WAN or its router
  • RTL8814AU WiFi chip USB 3.0 adapter connected to client PC via USB 3.0 ports
  • Mikrotik devices: regulatory domain TX power for 2.4 GHz, default manual-tx-power for 5 GHz
  • Default configuration for all devices: 2.4 GHz N-only, 5 GHz (auto)
  • Testing generated only 8.4 kB/s traffic
  • Stable link rates
  • No walls separating devices and test client
  • None of the test devices are Internet facing
  • Low interference environment, see spectral scan below. Scan not taken from test device but from the hAP ac outside of testing
  • Firmware: hAP ac2 (RouterOS + firmware: 6.43rc5), hAP ac (RouterOS + firmware: 6.41.3), Asus RT-N56U (30043807378)

Image
Last edited by squeeze on Wed May 16, 2018 2:14 pm, edited 5 times in total.
 
Sheriff1972
just joined
Posts: 20
Joined: Fri Feb 16, 2018 7:48 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Sun May 06, 2018 8:22 am

Amazing detective work! I hope that this gives a focus to the engineering team at MT.
 
hapi
Member Candidate
Member Candidate
Posts: 222
Joined: Fri Mar 11, 2011 11:21 am
Location: Czech Republic

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Sun May 06, 2018 12:45 pm

in 5GHz with "Intel Dual Band Wireless-AC 7265" is tx rate only 54Mbit. In 2.4GHz is full speed.

Image
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Sun May 06, 2018 3:43 pm

in 5GHz with "Intel Dual Band Wireless-AC 7265" is tx rate only 54Mbit. In 2.4GHz is full speed.

Image
Please consider to use the actual data transfer test tool, for instance iperf3.

Regarding to your Intel AC adapter, please select something link "boost performance" under the advanced tap of your wireless adapter in the device manage of the Windows PC.

YH
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 07, 2018 4:17 am

squeeze,

Reading your post in this thread gives more information about the current status of MikroTik WiFi products. As per my memory in other forum, that forum members concluded that MikroTik WiFi routers or APs would not their choices due to extremely poor performance. MikroTik products are good at routing and switching but WiFi. They will choose Ubiquiti APs to work with MikroTik routers or Ubiquiti routers. If we combine their comments and the results presented by you and morph1, it shows a very clear message that the overall performance of general MikroTik WiFi routers and APs is not satisfactory. I do not know whether the MikroTik Management Team knows this WiFi weakness. In order to resolve this poor Wifi performance, MikroTik have to form a good quality plan for their WiFi router and AP products by specifying the specifications for 1) ping test, 2) throughput test, and others against several brands of WiFi devices before they are released for production and sales.

Based on my past embedded firmware development's experience, it seems the interface between WiFi routines and RouterOS is not very good. It seems that the RouterOS do not have time to respond the request from the WiFi routine due to unknown reasons. Although the WiFi performance of hAP ac^2 has already been improved significantly as compared to the very beginning when the products launched to the market, it is still not enough as compared to the other WiFi brands.

I hope MikroTik has to review the interface between the RouterOS and WiFi routines throughly. If they can shorten the ping time, I do believe that the sales of MikroTik WiFi products will be increased tremendously.

YH
 
pepek
just joined
Posts: 23
Joined: Tue Apr 10, 2018 12:14 am
Location: CZ

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 10, 2018 6:08 pm

It seems, that there will be no speed improvement in near future because of other 6.42.1 problems being fixed in current RCs...
I think that Improving this problem to "somehow usable, not totally crappy" lowered it's priority

Still no response from support on my speed issues from last week...
 
startus
just joined
Posts: 2
Joined: Thu May 10, 2018 10:57 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 10, 2018 11:21 pm

hi everyone. I am sorry if this is not the right place to ask for an opinion, but I`ve decided to give a try once I`ve checked all the professional feedbacks about MK here and I felt it worths a try due to my outdated knowledge of MK products. My last and current version of RB is the 433AH that I use at home, but I want upgrade for a Giga + dual (2.4 and 5) wireless solution.

I almost bought the hAP ac^2, however after a quick search I`ve found this topic and basically I would like to have an idea if I should give a try and buy the hAP ac^2 (and wait for later improvements) or
buy the hAP AC RB962UiGS-5HacT2HnT model (I did not find a topic with problems for this model, is there any I don`t know?).

Both have the basic requirements I need and money is not a problem. Therefore I would like to know why not buy the hAP AC RB962UiGS-5HacT2HnT instead of the hAP ac^2

P.S. I just installed NAS with gigabit interface at home and I will upgrade the speed of my internet for something around 120mbs if that helps.

Thanks for the help.
 
Bivvy
newbie
Posts: 31
Joined: Sat Feb 04, 2017 1:36 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Fri May 11, 2018 8:21 am

Anyone know the first version of the firmware that suffered the poor WiFi problems?
We have >130 clients using HAP ac's and recently started getting complaints about poor speeds.
Bandwidth tests from the router and from PC's plugged straight into the router are all fine - it's just the WiFi that is letting us down.
As the problems started fairly recently and we've just gone through a process of upgrading all firmware to the latest stable release it seems like too much of a coincidence.
Now considering downgrading all firmware to a previous release - just need to know which one to go for...
Thanks
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Fri May 11, 2018 1:25 pm

Bivvy,

First of all, I guess you are talking about hAP ac^2 instead of hAP ac. They are two different models. Please note that the stable version of RouterOS on hAP ac^2 can not resolve the WiFi issues, you have to upgrade to the release candidate version of v6.43rc5 or above. Please update it to the current firmware of v6.43rc11 and you will find the speed improved significantly.

If you are not talking baout hAP ac^2, there will be other issues associated with the poor performance.

YH
 
yottabit
Member Candidate
Member Candidate
Posts: 160
Joined: Thu Feb 21, 2013 5:56 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Fri May 11, 2018 2:32 pm

Startus: the hAP ac model has superior wireless over the hAP ac2, with higher radio power and 3 antenna chains per radio instead of only 2.

But the hAP ac2 has a much more powerful CPU, quad-core ARM v. single-core MIPS.

I am using the ac2 for my router as the ac maxes CPU for me with my QoS rules. If you're going to use QoS and need NAT routing throughput greater than 100 Mbps, I suggest you use the ac2 for routing. And I do have three ac models that I use for wireless coverage and speed (I have the wireless function disabled on my ac2).

Sent from my Pixel XL using Tapatalk


 
startus
just joined
Posts: 2
Joined: Thu May 10, 2018 10:57 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Fri May 11, 2018 4:36 pm

yottabit: Thank you.

Just to clarify my scenario. Most of the time I will be using my desktop + NAS in the LAN side (gigabit) alongside with 5 wireless devices (2 cell phones, 1 tablet, 1 chromecast, and 1 raspberry pi).

From time to time (maybe once in a month) I may use the LAN to make backup`s of my NAS and local video streaming from the NAS, but not constantly.

I will use in the wan side (120mbs down and 20mbs up) with few rules, portforwarding and one VPN session. I also will enable some services in the NAS to be used from the internet, like audio streaming and file sharing (cloud) etc behind the NAT.

I don`t believe I will upgrade my internet in the next 3 years, being said, If I understood correctly, the hAP would do the job right? I think the hAP ac2 can help me as well even with less wireless power. I am afraid the problem with hAP ac2 may not be solved.

Anybody else could share any thoughts?
Thank you in advance.
 
yottabit
Member Candidate
Member Candidate
Posts: 160
Joined: Thu Feb 21, 2013 5:56 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Fri May 11, 2018 6:24 pm

Since your upstream speed is so low, you'll be just fine using the hAP ac as a router, even if you implement QoS in the future.

If you ever get 100M+ upstream, *and* want QoS, you can buy a more powerful router for that when you need it (e.g., hAP ac2, or whatever its successor may be).

Sent from my Pixel XL using Tapatalk

 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 14, 2018 3:51 am

startus,

Both hAP ac and hAP ac^2 can fit your requirements. As the member "yottabit" has mentioned that the main benefits of using hAP ac are the higher WiFi transmission power and throughput. The downside of hAP ac is the routing power.

Since both models can fit your requirement, you may consider to choose the lowest-priced one, that is hAP ac^2.

YH
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 14, 2018 2:05 pm

@startus

With release candidate RouterOS software (6.43rc5+) the hAP ac2 WiFi is good enough for anything for a typical consumer - though I would not recommend its WiFi for low latency applications at the moment, e.g. gaming, but you should be using wired for that anyway in most situations.

Since your main use case is less WiFi than strong routing, this actually makes the hAP ac2 absolutely preferable over the hAP ac or any other Mikrotik SOHO device.
 
Bivvy
newbie
Posts: 31
Joined: Sat Feb 04, 2017 1:36 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 14, 2018 6:50 pm

@yottabit - Thanks for your reply.
We're using HAP ac's (RB962UiGS-5HacT2HnT) - until you pointed out the difference I had not realised there were two versions. Sorry if I posted in the wrong place,
In the past fortnight we've had lots of complaints
a) slow WiFi performance
b) unable to connect over the WiFi

Downgrading from 6.42.1 to 6.40.4 seemed to fix the problems, but now I'm not so convinced as one of our customers is still struggling to connect to the WiFi even after the downgrade.
I appreciate that these issues may not be directly related to the original subject of this thread.

I like Mikrotik but can't afford to spend so much effort supporting the client routers...
Ticket raised with Mikrotik tech support but no response yet (although we're still inside the 3 day SLA)

Should I try the latest release candidate?
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 14, 2018 6:57 pm

Release candidates are not for production in any form of business environment. Nor are they intended as a substitute for basic troubleshooting.

Since your posts have nothing whatsover to do with the hAP ac2 (it is not even similar architecture, let alone radios), please refrain from bumping this topic. Many people are monitoring this thread for updates and information exclusive to the hAP ac2 and also receive notifications via email. Please respect them and start your own separate thread.

Thank you.
 
aidan
newbie
Posts: 28
Joined: Thu Jun 25, 2015 12:48 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Mon May 21, 2018 3:19 pm

Early reports mentioned that the hAP ac^2 and cAP AC have a 10 dBm weaker signal than a wAP AC. Can anyone comment whether 6.43rc has also resolved this problem? Thank you.
 
alengrosevic
just joined
Posts: 10
Joined: Fri Mar 30, 2018 12:14 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Wed May 23, 2018 11:34 pm

RC17 is out.
*) wireless - increased stability on hAP ac^2 and cAP ac with legacy data rates;

But after upgrade I cant login thru winbox anymore. It keeps saying wrong username password. Over Web page I can login, but over winbox I cant. Strange...

EDIT: Log from device:

Time May/23/2018 22:30:04
Buffer memory
Topics
system
error
critical
Message login failure for user admin from A4:**:**:**:**:B1 via winbox
 
mistry7
Forum Guru
Forum Guru
Posts: 1244
Joined: Tue Oct 13, 2009 11:57 am
Location: Germany

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Wed May 23, 2018 11:48 pm

RC17 is out.
*) wireless - increased stability on hAP ac^2 and cAP ac with legacy data rates;

But after upgrade I cant login thru winbox anymore. It keeps saying wrong username password. Over Web page I can login, but over winbox I cant. Strange...

EDIT: Log from device:

Time May/23/2018 22:30:04
Buffer memory
Topics
system
error
critical
Message login failure for user admin from A4:**:**:**:**:B1 via winbox
Could be this

winbox - improved authentication process excluding man-in-the-middle possibility (Winbox v3.14 required);

Or this:

user - all passwords are now hashed and encrypted, plaintext passwords are kept for downgrade (will be removed in later upgrades);
 
squeeze
Member Candidate
Member Candidate
Posts: 146
Joined: Thu Mar 22, 2018 7:53 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 24, 2018 4:52 am

*) wireless - increased stability on hAP ac^2 and cAP ac with legacy data rates;

I am going to wait until something as fundamental as logging in works, then I'll rerun the stability tests. How exciting, I didn't expect Mikrotik to update so fast on these ARM devices!
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 24, 2018 6:29 am

I have the same login problem as the others have. However I could login the router using the old version WinBox 2.2.18.

Based on my initial test, the ping time for 5GHz seems better than the previous versions.
 
alengrosevic
just joined
Posts: 10
Joined: Fri Mar 30, 2018 12:14 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 24, 2018 8:31 am

latest available winbox for download is 3.13. I cant find 3.14.

EDIT: Thank you guys for the link! :)
Last edited by alengrosevic on Thu May 24, 2018 1:21 pm, edited 2 times in total.
 
Zulu
just joined
Posts: 1
Joined: Thu May 24, 2018 8:38 am

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 24, 2018 8:41 am

latest available winbox for download is 3.13. I cant find 3.14.
new beta winbox https://www.mikrotik.com/download/share/winbox.exe
 
yhfung
Member Candidate
Member Candidate
Topic Author
Posts: 142
Joined: Tue Nov 20, 2012 6:58 pm

Re: hAP ac^2 Problems---Extremely Poor Performance found in 2.4G and 5G WiFi

Thu May 24, 2018 8:54 am

The line is at the bottom of the release announcement of v6.43rc17.

viewtopic.php?f=21&t=133420&start=100#p663507

In order to access router running this version by Winbox, please download v3.14rc Winbox loader from here:
https://www.mikrotik.com/download/share/winbox.exe

Who is online

Users browsing this forum: No registered users and 10 guests