Hi
following another thread that was closed already I am reopening this case because my issue wasn't solved yet.
I have hEX s as CAPsMAN and hAP ac^2 as AP
I have few lost of connectivity's and losing pings during some time to time - losing pings for 6-7 and once its back I see it reach 2000ms
I am exporting here the configuration I use
when I move the hAP ac^2 to be WISP AP as a standalone the pefromance is much much better. not the best as for my Meraki AP
on Meraki I could get full of 200mbit/200mbit connection, on hAP ac^2 I get at most ~150mbit .
attached both configuration files
any help will be appreciated.
Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
You do not have the required permissions to view the files attached to this post.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Enable local forwarding
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Its enabled on datapath1
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Use fixed channels.
Set all configuration (like country code, installation, etc.)
Nothing wrong with using DFS (only discomfort is the time it might take to scan)
With the settings supplied in your other topic I am able to get 400 ~ 450 Mbps (on the 5GHz radio of course).
Set all configuration (like country code, installation, etc.)
Nothing wrong with using DFS (only discomfort is the time it might take to scan)
With the settings supplied in your other topic I am able to get 400 ~ 450 Mbps (on the 5GHz radio of course).
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Hi
i found on https://bgp.guru/2020/01/19/centrally-m ... k-capsman/ that it claims that It needs to configure adaptive-nosie-immunity
could it be the case?
/interface wireless
set [ find default-name=wlan1 ] adaptive-noise-immunity=ap-and-client-mode
set [ find default-name=wlan2 ] adaptive-noise-immunity=ap-and-client-mode
I used fixed freq on the settings and filled out the country etc
also on your i found that there is a setting of hw retries 3 so I changed it as well
I will update with the findings so far
i found on https://bgp.guru/2020/01/19/centrally-m ... k-capsman/ that it claims that It needs to configure adaptive-nosie-immunity
could it be the case?
/interface wireless
set [ find default-name=wlan1 ] adaptive-noise-immunity=ap-and-client-mode
set [ find default-name=wlan2 ] adaptive-noise-immunity=ap-and-client-mode
I used fixed freq on the settings and filled out the country etc
also on your i found that there is a setting of hw retries 3 so I changed it as well
I will update with the findings so far
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
As I understand then you are pinging from wifi client device that is connected to AP? Try isolate issue by testing network stability from CAP to CAPSMAN server.
Last edited by maigonis on Sat Aug 08, 2020 8:16 pm, edited 1 time in total.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
But ac2 is connected via ether1 poe to the capsman (hex s)
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Can be port flapping.But ac2 is connected via ether1 poe to the capsman (hex s)
Check log and port status tab on both devices.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Couldn't find a flapping port. I have reset both devices to default. Set ac2 as cap and hex s as capsman and issue was reappear even when i connected the meraki so i am quite convinced that the issue is on hEX s . For testings i want to give few says with the ac^2 as main router and wifi as router.
I also noticed that ether2 was configured as the gateway ip and not as bridge that had the default ip 192.168.88.1 i am thinking that it could be bridge configuration mismatch.
I also noticed that ether2 was configured as the gateway ip and not as bridge that had the default ip 192.168.88.1 i am thinking that it could be bridge configuration mismatch.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Hi, usual mistakes .... ethernet interfaces that are ports to the bridge, are slave interfaces. Settings for them must be done on the bridge not on the interface.
- Set IP address on the bridge bridge1, not on ether2
- Make bridge1 member of the LAN interface list (ether2-ether5 are part of bridge1). If the ether2-5 are in the LAN list or not, that information does not harm, but is not used. Lucky as ether1 on the WAN list is wrong, if it is on the bridge also.
You tweak for performance (hwretries=3) but all performance tuning should be well understood. HWretries=3 will reduce the interface rate faster, and will disconnect faster, when there are retransmits.
Don't use a/n/ac if you do not need a, use n/ac . If you do not need n , use ac-only (=less overhead, better throughput)
Never rely on "auto" frequency with MKT. Set it yourselves after checking with Scan/Freq usage/Snooper for others to be avoided. Set Ceee or eCee,eeCe, eeeC , not XXXX as you don't know what will be choosen.
- Set IP address on the bridge bridge1, not on ether2
- Make bridge1 member of the LAN interface list (ether2-ether5 are part of bridge1). If the ether2-5 are in the LAN list or not, that information does not harm, but is not used. Lucky as ether1 on the WAN list is wrong, if it is on the bridge also.
You tweak for performance (hwretries=3) but all performance tuning should be well understood. HWretries=3 will reduce the interface rate faster, and will disconnect faster, when there are retransmits.
Don't use a/n/ac if you do not need a, use n/ac . If you do not need n , use ac-only (=less overhead, better throughput)
Never rely on "auto" frequency with MKT. Set it yourselves after checking with Scan/Freq usage/Snooper for others to be avoided. Set Ceee or eCee,eeCe, eeeC , not XXXX as you don't know what will be choosen.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Ether2 was set as 192.168.88.1 probably as default and not by me.
Following what you said about slaves it means that i dont need to configure bridge for CAP?
When ether1 on ac2 was on a bridge it was when i tried to power up the ac2 with poe in. I figured out that hex s doesnt have enough power to do it properly eventhough it was working.
Who could tell what capsman configuration are recommended to avoid any lose or cut of bandwidth. And also what are the rates being recommended to be use.
Tnx
Following what you said about slaves it means that i dont need to configure bridge for CAP?
When ether1 on ac2 was on a bridge it was when i tried to power up the ac2 with poe in. I figured out that hex s doesnt have enough power to do it properly eventhough it was working.
Who could tell what capsman configuration are recommended to avoid any lose or cut of bandwidth. And also what are the rates being recommended to be use.
Tnx
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
BS......... you control the config, so whether it was set or not its your decision to keep it or leave it...............Ether2 was set as 192.168.88.1 probably as default and not by me.
-
-
nescafe2002
Forum Veteran
- Posts: 915
- Joined:
- Location: Netherlands
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Actually 192.168.88.1 on (slave interface) ether2 is a default configuration (or quick set) thingy. And; I think we're here to share & learn, not to call BS on each other.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
... on bridge ... says the hAPac2 ... 
You do not have the required permissions to view the files attached to this post.
-
-
nescafe2002
Forum Veteran
- Posts: 915
- Joined:
- Location: Netherlands
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
... or quick set (after checking "Bridge All LAN Ports"):
You do not have the required permissions to view the files attached to this post.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
well well ... going back to Quick Set after doing anything in Winbox or webfig ... is a guaranteed way to get in trouble.
Never ever re-enter Quick-Set if something has been touched. The scripts are not smart, and do not consider whatever someone has done except what was in the default config.
Would not be surprised there are indeed mis-configs in Quick Set. After some experience one stays away from Quick Set.
There are 7 Quick Sets for the hAP ac2. And which is what? What if you change Quick Set choice while entering data? These Quick Sets are not the Default Config.
"ethernet" is not one of the hApac2 Quick Sets, you are on CHR.
Never ever re-enter Quick-Set if something has been touched. The scripts are not smart, and do not consider whatever someone has done except what was in the default config.
Would not be surprised there are indeed mis-configs in Quick Set. After some experience one stays away from Quick Set.
There are 7 Quick Sets for the hAP ac2. And which is what? What if you change Quick Set choice while entering data? These Quick Sets are not the Default Config.
"ethernet" is not one of the hApac2 Quick Sets, you are on CHR.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
I've reset all to defaults
Who could suggest which rates settings I should use?
Who could suggest which rates settings I should use?
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Leave the rates on default. Wifi is dynamically adjusting. The wifi performance can be followed in wireless tables menu, tab "registrations".
In tab list check signal level , TX/RX CCQ, and interface rate used (rate/bandwidth/streams/guard).
This is essential information before touching anything in the wireless settings.
CCQ values may be low until there is enough traffic generated.
In tab list check signal level , TX/RX CCQ, and interface rate used (rate/bandwidth/streams/guard).
This is essential information before touching anything in the wireless settings.
CCQ values may be low until there is enough traffic generated.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
When you mention defaults means not to config ratesin capsman?
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
always getting slow on download speed and good on upload
https://www.speedtest.net/result/9992998359
how come?
https://www.speedtest.net/result/9992998359
how come?
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Sorry, I configure directly on the AP. Those defaults are meaningful. I see similar defaults in CAPsMAN in the wiki, except for 802.11ac.
Checking the wiki ....
rates.basic (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6Mbps and maybe also 9Mbps (error in wiki, it's 9 not 11) and 12Mbps, but those later 2 is tuning already.
rates.supported (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6,9,12,18,24,36,48,54Mbps We don't want/need the b range 1,2,5.5,11Mbps. It's used for 802.11a and 802.11g.
rates.ht-basic-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7) ----> yes good range of defaults = single stream all n rates possible
rates.ht-supported-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7; mcs-8; mcs-9; mcs-10; mcs-11; mcs-12; mcs-13; mcs-14; mcs-15; mcs-16; mcs-17; mcs-18; mcs-19; mcs-20; mcs-21; mcs-22; mcs-23) -----> good range of defaults = all n rates for up to 3 streams/chains/radios
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - client must support MCS-0 to MCS-7
MCS 0-8 - client must support MCS-0 to MCS-8
MCS 0-9 - client must support MCS-0 to MCS-9
rates.vht-supported-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -----> really "none" ???? default=none => disables 802.11ac . none=disable !. Good default is, expecting a max of 3 radios/streams/chains : "MCS 0-9, MCS 0-9, MCS 0-9" as is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - devices will advertise as supported MCS-0 to MCS-7
MCS 0-8 - devices will advertise as supported MCS-0 to MCS-8
MCS 0-9 - devices will advertise as supported MCS-0 to MCS-9
Checking the wiki ....
rates.basic (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6Mbps and maybe also 9Mbps (error in wiki, it's 9 not 11) and 12Mbps, but those later 2 is tuning already.
rates.supported (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6,9,12,18,24,36,48,54Mbps We don't want/need the b range 1,2,5.5,11Mbps. It's used for 802.11a and 802.11g.
rates.ht-basic-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7) ----> yes good range of defaults = single stream all n rates possible
rates.ht-supported-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7; mcs-8; mcs-9; mcs-10; mcs-11; mcs-12; mcs-13; mcs-14; mcs-15; mcs-16; mcs-17; mcs-18; mcs-19; mcs-20; mcs-21; mcs-22; mcs-23) -----> good range of defaults = all n rates for up to 3 streams/chains/radios
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - client must support MCS-0 to MCS-7
MCS 0-8 - client must support MCS-0 to MCS-8
MCS 0-9 - client must support MCS-0 to MCS-9
rates.vht-supported-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -----> really "none" ???? default=none => disables 802.11ac . none=disable !. Good default is, expecting a max of 3 radios/streams/chains : "MCS 0-9, MCS 0-9, MCS 0-9" as is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - devices will advertise as supported MCS-0 to MCS-7
MCS 0-8 - devices will advertise as supported MCS-0 to MCS-8
MCS 0-9 - devices will advertise as supported MCS-0 to MCS-9
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
What are the values in your client registration? TX rate, RX rate, TX/RX CCQ, Signal strength, .....always getting slow on download speed and good on upload
https://www.speedtest.net/result/9992998359
how come?
May be something else than the wiki link also !
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
What are the values in your client registration? TX rate, RX rate, TX/RX CCQ, Signal strength, .....always getting slow on download speed and good on upload
https://www.speedtest.net/result/9992998359
how come?
May be something else than the wiki link also !
See attached
You do not have the required permissions to view the files attached to this post.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
What are the default rates in capsman if rates are not configured?Sorry, I configure directly on the AP. Those defaults are meaningful. I see similar defaults in CAPsMAN in the wiki, except for 802.11ac.
Checking the wiki ....
rates.basic (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6Mbps and maybe also 9Mbps (error in wiki, it's 9 not 11) and 12Mbps, but those later 2 is tuning already.
rates.supported (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6,9,12,18,24,36,48,54Mbps We don't want/need the b range 1,2,5.5,11Mbps. It's used for 802.11a and 802.11g.
rates.ht-basic-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7) ----> yes good range of defaults = single stream all n rates possible
rates.ht-supported-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7; mcs-8; mcs-9; mcs-10; mcs-11; mcs-12; mcs-13; mcs-14; mcs-15; mcs-16; mcs-17; mcs-18; mcs-19; mcs-20; mcs-21; mcs-22; mcs-23) -----> good range of defaults = all n rates for up to 3 streams/chains/radios
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - client must support MCS-0 to MCS-7
MCS 0-8 - client must support MCS-0 to MCS-8
MCS 0-9 - client must support MCS-0 to MCS-9
rates.vht-supported-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -----> really "none" ???? default=none => disables 802.11ac . none=disable !. Good default is, expecting a max of 3 radios/streams/chains : "MCS 0-9, MCS 0-9, MCS 0-9" as is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - devices will advertise as supported MCS-0 to MCS-7
MCS 0-8 - devices will advertise as supported MCS-0 to MCS-8
MCS 0-9 - devices will advertise as supported MCS-0 to MCS-9
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
see BOLD from the wiki copy above.What are the default rates in capsman if rates are not configured?Sorry, I configure directly on the AP. Those defaults are meaningful. I see similar defaults in CAPsMAN in the wiki, except for 802.11ac.
Checking the wiki ....
rates.basic (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6Mbps and maybe also 9Mbps (error in wiki, it's 9 not 11) and 12Mbps, but those later 2 is tuning already.
rates.supported (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6,9,12,18,24,36,48,54Mbps We don't want/need the b range 1,2,5.5,11Mbps. It's used for 802.11a and 802.11g.
rates.ht-basic-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7) ----> yes good range of defaults = single stream all n rates possible
rates.ht-supported-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7; mcs-8; mcs-9; mcs-10; mcs-11; mcs-12; mcs-13; mcs-14; mcs-15; mcs-16; mcs-17; mcs-18; mcs-19; mcs-20; mcs-21; mcs-22; mcs-23) -----> good range of defaults = all n rates for up to 3 streams/chains/radios
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - client must support MCS-0 to MCS-7
MCS 0-8 - client must support MCS-0 to MCS-8
MCS 0-9 - client must support MCS-0 to MCS-9
rates.vht-supported-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -----> really "none" ???? default=none => disables 802.11ac . none=disable !. Good default is, expecting a max of 3 radios/streams/chains : "MCS 0-9, MCS 0-9, MCS 0-9" as is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - devices will advertise as supported MCS-0 to MCS-7
MCS 0-8 - devices will advertise as supported MCS-0 to MCS-8
MCS 0-9 - devices will advertise as supported MCS-0 to MCS-9
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
this is how i have configured it
is it ok?both for 5G and 2.4G
Code: Select all
/caps-man rates
add name=default-rate basic=6Mbps supported=6Mbps,9Mbps,12Mbps,18Mbps,24Mbps,36Mbps,48Mbps,54Mbps ht-basic-mcs=mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7 ht-supported-mcs=mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15,mcs-16,mcs-17,mcs-18,mcs-19,mcs-20,mcs-21,mcs-22,mcs-23 vht-basic-mcs=mcs0-7 vht-supported-mcs=mcs0-9,mcs0-9,mcs0-9
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Interesting registration.
Signal -55 dBm: is OK
TX rate 292,5Mbps-80MHz/1S .... you get MCS7/ single stream/long guard (http://mcsindex.com/) . Expected net data rate single direction is 146Mbps
RX rate 433Mbps-80MHz/1S/SGI you get MCS9 (best possible)/single stream/short guard. That's the best you could ever get from 80MHz single stream. Expected net data rate single direction is 215 Mbps.
TX rate set : (most interesting) You do use VHT MCS 0-9 on a single stream !?! what makes it report a single stream/chain/radio ?!? It is at least not the "none" from the wiki. but is not dual stream neither.as reported here. The hAPac2 has dual stream. What about the client device? Dual stream would double the throughput.
Why is this TX rate lower?? You will probably see a TX/RX CCQ around 90/100 % . This means the signal from the hAP ac2 is disturbed at the client receiver point. Therefor 802.11 lowers the rate and the encoding, to get it through. Why is the signal disturbed? Well that's looking for reflections on wall's and metal or other devices around the client and for obstructions in the path.
Signal -55 dBm: is OK
TX rate 292,5Mbps-80MHz/1S .... you get MCS7/ single stream/long guard (http://mcsindex.com/) . Expected net data rate single direction is 146Mbps
RX rate 433Mbps-80MHz/1S/SGI you get MCS9 (best possible)/single stream/short guard. That's the best you could ever get from 80MHz single stream. Expected net data rate single direction is 215 Mbps.
TX rate set : (most interesting) You do use VHT MCS 0-9 on a single stream !?! what makes it report a single stream/chain/radio ?!? It is at least not the "none" from the wiki. but is not dual stream neither.as reported here. The hAPac2 has dual stream. What about the client device? Dual stream would double the throughput.
Why is this TX rate lower?? You will probably see a TX/RX CCQ around 90/100 % . This means the signal from the hAP ac2 is disturbed at the client receiver point. Therefor 802.11 lowers the rate and the encoding, to get it through. Why is the signal disturbed? Well that's looking for reflections on wall's and metal or other devices around the client and for obstructions in the path.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
It was direct near the access point
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
I dont understand why download stream is so low compares to the up stream
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Sorry, I configure directly on the AP. Those defaults are meaningful. I see similar defaults in CAPsMAN in the wiki, except for 802.11ac.
Checking the wiki ....
rates.basic (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6Mbps and maybe also 9Mbps (error in wiki, it's 9 not 11) and 12Mbps, but those later 2 is tuning already.
rates.supported (1Mbps | 2Mbps | 5.5Mbps | 6Mbps | 11Mbps | 11Mbps | 12Mbps | 18Mbps | 24Mbps | 36Mbps | 48Mbps | 54Mbps; Default: ) -> what is default here ????????? Should be 6,9,12,18,24,36,48,54Mbps We don't want/need the b range 1,2,5.5,11Mbps. It's used for 802.11a and 802.11g.
rates.ht-basic-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7) ----> yes good range of defaults = single stream all n rates possible
rates.ht-supported-mcs (list of (mcs-0 | mcs-1 | mcs-2 | mcs-3 | mcs-4 | mcs-5 | mcs-6 | mcs-7 | mcs-8 | mcs-9 | mcs-10 | mcs-11 | mcs-12 | mcs-13 | mcs-14 | mcs-15 | mcs-16 | mcs-17 | mcs-18 | mcs-19 | mcs-20 | mcs-21 | mcs-22 | mcs-23); Default: mcs-0; mcs-1; mcs-2; mcs-3; mcs-4; mcs-5; mcs-6; mcs-7; mcs-8; mcs-9; mcs-10; mcs-11; mcs-12; mcs-13; mcs-14; mcs-15; mcs-16; mcs-17; mcs-18; mcs-19; mcs-20; mcs-21; mcs-22; mcs-23) -----> good range of defaults = all n rates for up to 3 streams/chains/radios
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - client must support MCS-0 to MCS-7
MCS 0-8 - client must support MCS-0 to MCS-8
MCS 0-9 - client must support MCS-0 to MCS-9
rates.vht-supported-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -----> really "none" ???? default=none => disables 802.11ac . none=disable !. Good default is, expecting a max of 3 radios/streams/chains : "MCS 0-9, MCS 0-9, MCS 0-9" as is set in the AP as default.
none - will not use selected Spatial Stream
MCS 0-7 - devices will advertise as supported MCS-0 to MCS-7
MCS 0-8 - devices will advertise as supported MCS-0 to MCS-8
MCS 0-9 - devices will advertise as supported MCS-0 to MCS-9
How do you configure it directly on the ap if the ap is a CAP?
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
You are close to your AP. The client device normally sends a weaker signal than an AP. The client device signal is already received at -55 dBm, and the client is probably sending at 10dBm. The AP may be sending at 20 to even 27 dBm. This can be too strong for the room and distance you are in, leading to massive reflections and distortions. Lower the TX power of the AP to test this theory. Set it on something like 11 dBm.
Another reason is that you measure the whole internet chain. Not the local wifi speed. Local wifi speed is expected to be 146Mbps. But you get lower internet speed with speedtest. Reasons could be in the TCP congestion control algorithm used (https://en.wikipedia.org/wiki/TCP_congestion_control ) . Other reason could be that you are doing other background downloads at the same time. The only direct check value is the "registration" reading. It's a live indicator. Do one step away from the AP and see the difference immediately. (It needs some traffic to update properly). Netspeed has too many other influencing things, to be usable as differentiating diagnostic tool.
If you get your wifi TX rate at 433Mbps and there is still a difference between up and download ..... it's not your wifi that is the bottleneck.
I hope you removed the hw-retries=3 from your config, leave it on default=7. You do not want wifi to drop packets, it might trigger TCP congestion control.
https://www.google.com/search?q=why+is+ ... +my+upload
Another reason is that you measure the whole internet chain. Not the local wifi speed. Local wifi speed is expected to be 146Mbps. But you get lower internet speed with speedtest. Reasons could be in the TCP congestion control algorithm used (https://en.wikipedia.org/wiki/TCP_congestion_control ) . Other reason could be that you are doing other background downloads at the same time. The only direct check value is the "registration" reading. It's a live indicator. Do one step away from the AP and see the difference immediately. (It needs some traffic to update properly). Netspeed has too many other influencing things, to be usable as differentiating diagnostic tool.
If you get your wifi TX rate at 433Mbps and there is still a difference between up and download ..... it's not your wifi that is the bottleneck.
I hope you removed the hw-retries=3 from your config, leave it on default=7. You do not want wifi to drop packets, it might trigger TCP congestion control.
https://www.google.com/search?q=why+is+ ... +my+upload
Last edited by bpwl on Sun Aug 30, 2020 10:36 am, edited 1 time in total.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
[/quote]
How do you configure it directly on the ap if the ap is a CAP?
[/quote]
Take your hAP ac2. Standard config. Connect via wifi or ether2-5. Remove the DHCP server from the bridge, move the DHCP client from ether1 to the bridge. Add ether1 to the bridge.
All ether ports and the wifi's are now in the LAN interface list, they form one LAN. Firewall and NAT is not used as you bridge everything, and nothing is really (as master) in the WAN interface list.
How do you configure it directly on the ap if the ap is a CAP?
[/quote]
Take your hAP ac2. Standard config. Connect via wifi or ether2-5. Remove the DHCP server from the bridge, move the DHCP client from ether1 to the bridge. Add ether1 to the bridge.
All ether ports and the wifi's are now in the LAN interface list, they form one LAN. Firewall and NAT is not used as you bridge everything, and nothing is really (as master) in the WAN interface list.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
I want the ac2 to be cap so how is it realted to local config?
For the power should i set it for 5gh or 2.4?
Sent from my SM-A705FN using Tapatalk
For the power should i set it for 5gh or 2.4?
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
And i dont have hw retries configured.
Sent from my SM-A705FN using Tapatalk
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Sorry, was in another postAnd i dont have hw retries configured.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Somehow i have to set to none, when i set to mcs 0-7 it will cause some my older device (802.11n) unable to connect.
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Can you confirm that its ok?Somehow i have to set to none, when i set to mcs 0-7 it will cause some my older device (802.11n) unable to connect.
rates.vht-basic-mcs (none | MCS 0-7 | MCS 0-8 | MCS 0-9; Default: none)
You can set MCS interval for each of Spatial Stream -------> they must be kidding! This DISABLES 802.11ac mode of operation. Good default is "MCS0-7" like it is set in the AP as default.
viewtopic.php?p=813901#p813901
Sent from my SM-A705FN using Tapatalk
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
Strange! 802.11n is NOT using VHT MCS at all, only HT MCS. For your older device mode must be "n/ac" , not "ac-only".
Somehow i have to set to none, when i set to mcs 0-7 it will cause some my older device (802.11n) unable to connect.
However I do not know what effects a CAPsMAN setting has on the real setting of the interface. And the default of the wiki (=none), seems not to be the real default (= :, or not specified) !?!?
Last edited by bpwl on Sun Aug 30, 2020 11:56 pm, edited 2 times in total.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
@Shy, settings seems OK. But I don't use CAPsMAN so far , so I don't know what the ultimate setting on the interface will be. And with RouterOS I'm not sure what of that setting is actually used.
Re: Bad performance on hAP ac^2 vs CAPsMAN AP behind hEX s as AP
I switched to hAP ac² to be as standalone and performance is MUCH better than being CAP.