Fix Harry's wifi: Difference between revisions
(→Tools) |
|||
(24 intermediate revisions by the same user not shown) | |||
Line 3: | Line 3: | ||
So I'm practicing on our wifi and trying to make sure I've got all my tools together. These procedures assume that I have access to an Android phone and a linux laptop both on the wifi network of interest. | So I'm practicing on our wifi and trying to make sure I've got all my tools together. These procedures assume that I have access to an Android phone and a linux laptop both on the wifi network of interest. | ||
===Tools=== | |||
* Netalyzr http://netalyzr.icsi.berkeley.edu/ | |||
** Hotel wifi in Indiana http://n1.netalyzr.icsi.berkeley.edu/summary/id=43ca253f-14484-771d32b8-94ff-43ca-afa3#feedback | |||
** T-Mobile tethered hotspot http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-4364-e6932437-bd46-4a91-a8af | |||
** wired at home http://n1.netalyzr.icsi.berkeley.edu/summary/id=43ca253f-13285-09ce446b-2086-42c7-a70d | |||
** wireless at home http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-4068-6244943c-54c1-46b4-94fd | |||
** wireless at San Jose Marriott http://netalyzr.icsi.berkeley.edu/restore/id=36ea240d-4004-2662815a-ac01-4ff9-a91d | |||
** wireless at Harlan's http://n2.netalyzr.icsi.berkeley.edu/summary/id=43ca208a-6327-fe352067-5b70-4cfc-af4a | |||
** wireless at Newport Library http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-8858-a19ce6f9-cfd6-4f48-afc1 | |||
*** blocks ports, boo | |||
* Ntop http://www.linuxforu.com/2010/10/ntop-network-diagnostics-tool/ | |||
* Chrome connectivity diagnostics plugin https://chrome.google.com/webstore/detail/chrome-connectivity-diagn/eemlkeanncmjljgehlbplemhmdmalhdc?hl=en | |||
===Details=== | |||
====Android Wifi Analyzer for signal strength==== | ====Android Wifi Analyzer for signal strength==== | ||
Line 57: | Line 72: | ||
* across portlandia ping time between a desktop and a tablet is between 1.9 and 99ms. Average is 57ms. | * across portlandia ping time between a desktop and a tablet is between 1.9 and 99ms. Average is 57ms. | ||
* there is no measurable packet loss | * there is no measurable packet loss | ||
====use firebug net to measure stages of web page load==== | |||
{| border="1" cellspacing="0" cellpadding="5" align="center" | |||
!Location | |||
!DNS ms | |||
!Connecting ms | |||
!Sending ms | |||
!Waiting ms | |||
!Receiving ms | |||
!Notes | |||
|- | |||
|Home wired | |||
|35 | |||
|0 | |||
|0 | |||
|228 | |||
|56 | |||
| | |||
|- | |||
|Harlans | |||
|0 | |||
|120 | |||
|0 | |||
|307 | |||
|174 | |||
|Used Chrome. Funny that it doesn't show slow DNS. | |||
|} | |||
====how many hops to DNS?==== | ====how many hops to DNS?==== | ||
At home I have 14 hops. | |||
<pre> | |||
root@ferret:~# mtr --report 8.8.8.8 | |||
Start: Tue Feb 4 18:09:22 2014 | |||
HOST: ferret Loss% Snt Last Avg Best Wrst StDev | |||
1.|-- weasel.local 0.0% 10 0.1 0.1 0.1 0.2 0.0 | |||
2.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 | |||
3.|-- xe-0-0-3-sur03.troutdale. 0.0% 10 8.1 8.0 4.9 13.7 2.2 | |||
4.|-- ae-53-0-ar03.beaverton.or 0.0% 10 9.2 18.3 7.6 69.6 20.3 | |||
5.|-- ae-1-0-ar03.troutdale.or. 0.0% 10 7.5 13.7 7.5 43.6 11.6 | |||
6.|-- he-2-0-0-0-11-cr01.seattl 0.0% 10 13.3 13.7 11.6 15.2 1.1 | |||
7.|-- be-10-pe03.seattle.wa.ibo 0.0% 10 15.1 13.1 12.0 15.1 0.7 | |||
8.|-- as15169-1-c.seattle.wa.ib 0.0% 10 17.8 13.7 12.1 17.8 1.7 | |||
9.|-- 209.85.249.32 0.0% 10 11.6 12.0 10.9 13.1 0.3 | |||
10.|-- 66.249.94.199 0.0% 10 12.6 13.1 11.9 15.4 1.1 | |||
11.|-- 216.239.46.200 0.0% 10 20.0 19.5 18.4 21.6 0.8 | |||
12.|-- 216.239.48.167 0.0% 10 20.7 19.4 15.1 22.8 1.9 | |||
13.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 | |||
14.|-- google-public-dns-a.googl 0.0% 10 19.1 19.5 15.9 22.7 1.9 | |||
</pre> | |||
At Harlan's there are 11 hops, but it is slower. Almost 100ms just to get to DNS. | |||
<pre> | |||
[root@servo rday]# mtr --report 8.8.8.8 | |||
Start: Fri May 2 16:10:55 2014 | |||
HOST: servo.finninday.net Loss% Snt Last Avg Best Wrst StDev | |||
1.|-- dslmodem.domain 0.0% 10 2.1 1.9 1.7 2.1 0.0 | |||
2.|-- lactwahed11 0.0% 10 30.7 31.8 30.7 34.4 1.1 | |||
3.|-- lsancarccor52 0.0% 10 54.1 52.8 51.6 54.1 0.6 | |||
4.|-- lsancarccor52 0.0% 10 52.8 52.7 51.9 53.1 0.0 | |||
5.|-- 72.14.234.45 70.0% 10 53.9 60.0 53.9 72.3 10.6 | |||
6.|-- 64.233.174.188 0.0% 10 53.0 53.4 52.2 55.8 0.9 | |||
7.|-- 216.239.46.154 0.0% 10 102.2 100.1 95.4 111.6 6.0 | |||
8.|-- 216.239.46.55 0.0% 10 95.8 95.5 94.2 97.9 0.7 | |||
9.|-- 216.239.46.191 0.0% 10 95.1 98.7 94.1 132.4 11.8 | |||
10.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 | |||
11.|-- google 0.0% 10 94.4 94.9 94.4 95.7 0.0 | |||
</pre> | |||
====how fast is DNS?==== | ====how fast is DNS?==== | ||
I can expect a response in 20 to 35 ms. | I can expect a response in 20 to 35 ms. | ||
Line 83: | Line 164: | ||
;; WHEN: Tue Feb 04 18:07:56 PST 2014 | ;; WHEN: Tue Feb 04 18:07:56 PST 2014 | ||
;; MSG SIZE rcvd: 58 | ;; MSG SIZE rcvd: 58 | ||
</pre> | |||
Harlan DNS is slower. Over 100ms. | |||
<pre> | |||
[root@servo rday]# dig finninday.net | |||
; <<>> DiG 9.9.4-P2-RedHat-9.9.4-12.P2.fc20 <<>> finninday.net | |||
;; global options: +cmd | |||
;; Got answer: | |||
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 61312 | |||
;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 2, ADDITIONAL: 2 | |||
;; QUESTION SECTION: | |||
;finninday.net. IN A | |||
;; ANSWER SECTION: | |||
finninday.net. 410 IN A 216.99.216.99 | |||
;; AUTHORITY SECTION: | |||
finninday.net. 79345 IN NS yns2.yahoo.com. | |||
finninday.net. 79345 IN NS yns1.yahoo.com. | |||
;; ADDITIONAL SECTION: | |||
yns1.yahoo.com. 63133 IN A 67.195.1.92 | |||
yns2.yahoo.com. 63133 IN A 98.139.247.192 | |||
;; Query time: 117 msec | |||
;; SERVER: 192.168.0.1#53(192.168.0.1) | |||
;; WHEN: Fri May 02 16:09:15 PDT 2014 | |||
;; MSG SIZE rcvd: 126 | |||
</pre> | </pre> |
Latest revision as of 15:07, 30 September 2014
Background
A friend suffers from crappy wifi. I keep telling him that I'll fix it, but I haven't yet. We spend time talking or eating instead of poking at the wifi. Before I go over again, I want to have a streamlined process for diagnosing the problem with all the necessary tools pre-loaded on my laptop so I can look like a rock star.
So I'm practicing on our wifi and trying to make sure I've got all my tools together. These procedures assume that I have access to an Android phone and a linux laptop both on the wifi network of interest.
Tools
- Netalyzr http://netalyzr.icsi.berkeley.edu/
- Hotel wifi in Indiana http://n1.netalyzr.icsi.berkeley.edu/summary/id=43ca253f-14484-771d32b8-94ff-43ca-afa3#feedback
- T-Mobile tethered hotspot http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-4364-e6932437-bd46-4a91-a8af
- wired at home http://n1.netalyzr.icsi.berkeley.edu/summary/id=43ca253f-13285-09ce446b-2086-42c7-a70d
- wireless at home http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-4068-6244943c-54c1-46b4-94fd
- wireless at San Jose Marriott http://netalyzr.icsi.berkeley.edu/restore/id=36ea240d-4004-2662815a-ac01-4ff9-a91d
- wireless at Harlan's http://n2.netalyzr.icsi.berkeley.edu/summary/id=43ca208a-6327-fe352067-5b70-4cfc-af4a
- wireless at Newport Library http://n3.netalyzr.icsi.berkeley.edu/summary/id=36ea240d-8858-a19ce6f9-cfd6-4f48-afc1
- blocks ports, boo
- Ntop http://www.linuxforu.com/2010/10/ntop-network-diagnostics-tool/
- Chrome connectivity diagnostics plugin https://chrome.google.com/webstore/detail/chrome-connectivity-diagn/eemlkeanncmjljgehlbplemhmdmalhdc?hl=en
Details
Android Wifi Analyzer for signal strength
At home I get a maximum signal of about -50 db from Portlandia when the phone is less than 5 feet from it. I get about -75 db from ahub which is one floor down from the phone.
Identify protocol
Standard android wifi settings page tells me link speed, security type, and IP.
- Portlandia: 54Mbps (802.11g)
- ahub: 36Mbps (mixed 802.11b+g)
Other possibilities are:
- 802.11n (2009) up to 150Mbps
- 802.11ac (2012) up to 866Mbps
see what dhcp is giving us
root@ferret:~# less /var/lib/dhcp/dhclient.leases lease { interface "eth0"; fixed-address 10.0.0.6; option subnet-mask 255.0.0.0; option routers 10.0.0.3; option dhcp-lease-time 43200; option dhcp-message-type 5; option domain-name-servers 8.8.8.8,8.8.4.4; option dhcp-server-identifier 10.0.0.3; option domain-name "finninday.net"; renew 0 2013/11/24 04:12:33; rebind 0 2013/11/24 08:53:29; expire 0 2013/11/24 10:23:29; }
what does the routing table look like?
root@ferret:~# ip route list default via 10.0.0.3 dev eth0 proto static 10.0.0.0/8 dev eth0 proto kernel scope link src 10.0.0.6 metric 1 root@ferret:~# route -n Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 0.0.0.0 10.0.0.3 0.0.0.0 UG 0 0 0 eth0 10.0.0.0 0.0.0.0 255.0.0.0 U 1 0 0 eth0
measure ping time and packet loss to default gateway with mtr
- on a wired network ping time is between 0.2 and 0.4ms
- across portlandia ping time between a desktop and a tablet is between 1.9 and 99ms. Average is 57ms.
- there is no measurable packet loss
use firebug net to measure stages of web page load
Location | DNS ms | Connecting ms | Sending ms | Waiting ms | Receiving ms | Notes |
---|---|---|---|---|---|---|
Home wired | 35 | 0 | 0 | 228 | 56 | |
Harlans | 0 | 120 | 0 | 307 | 174 | Used Chrome. Funny that it doesn't show slow DNS. |
how many hops to DNS?
At home I have 14 hops.
root@ferret:~# mtr --report 8.8.8.8 Start: Tue Feb 4 18:09:22 2014 HOST: ferret Loss% Snt Last Avg Best Wrst StDev 1.|-- weasel.local 0.0% 10 0.1 0.1 0.1 0.2 0.0 2.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 3.|-- xe-0-0-3-sur03.troutdale. 0.0% 10 8.1 8.0 4.9 13.7 2.2 4.|-- ae-53-0-ar03.beaverton.or 0.0% 10 9.2 18.3 7.6 69.6 20.3 5.|-- ae-1-0-ar03.troutdale.or. 0.0% 10 7.5 13.7 7.5 43.6 11.6 6.|-- he-2-0-0-0-11-cr01.seattl 0.0% 10 13.3 13.7 11.6 15.2 1.1 7.|-- be-10-pe03.seattle.wa.ibo 0.0% 10 15.1 13.1 12.0 15.1 0.7 8.|-- as15169-1-c.seattle.wa.ib 0.0% 10 17.8 13.7 12.1 17.8 1.7 9.|-- 209.85.249.32 0.0% 10 11.6 12.0 10.9 13.1 0.3 10.|-- 66.249.94.199 0.0% 10 12.6 13.1 11.9 15.4 1.1 11.|-- 216.239.46.200 0.0% 10 20.0 19.5 18.4 21.6 0.8 12.|-- 216.239.48.167 0.0% 10 20.7 19.4 15.1 22.8 1.9 13.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 14.|-- google-public-dns-a.googl 0.0% 10 19.1 19.5 15.9 22.7 1.9
At Harlan's there are 11 hops, but it is slower. Almost 100ms just to get to DNS.
[root@servo rday]# mtr --report 8.8.8.8 Start: Fri May 2 16:10:55 2014 HOST: servo.finninday.net Loss% Snt Last Avg Best Wrst StDev 1.|-- dslmodem.domain 0.0% 10 2.1 1.9 1.7 2.1 0.0 2.|-- lactwahed11 0.0% 10 30.7 31.8 30.7 34.4 1.1 3.|-- lsancarccor52 0.0% 10 54.1 52.8 51.6 54.1 0.6 4.|-- lsancarccor52 0.0% 10 52.8 52.7 51.9 53.1 0.0 5.|-- 72.14.234.45 70.0% 10 53.9 60.0 53.9 72.3 10.6 6.|-- 64.233.174.188 0.0% 10 53.0 53.4 52.2 55.8 0.9 7.|-- 216.239.46.154 0.0% 10 102.2 100.1 95.4 111.6 6.0 8.|-- 216.239.46.55 0.0% 10 95.8 95.5 94.2 97.9 0.7 9.|-- 216.239.46.191 0.0% 10 95.1 98.7 94.1 132.4 11.8 10.|-- ??? 100.0 10 0.0 0.0 0.0 0.0 0.0 11.|-- google 0.0% 10 94.4 94.9 94.4 95.7 0.0
how fast is DNS?
I can expect a response in 20 to 35 ms.
root@ferret:~# dig finninday.net ; <<>> DiG 9.9.3-rpz2+rl.13214.22-P2-Ubuntu-1:9.9.3.dfsg.P2-4ubuntu1.1 <<>> finninday.net ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 28458 ;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 512 ;; QUESTION SECTION: ;finninday.net. IN A ;; ANSWER SECTION: finninday.net. 217 IN A 216.99.216.99 ;; Query time: 20 msec ;; SERVER: 8.8.8.8#53(8.8.8.8) ;; WHEN: Tue Feb 04 18:07:56 PST 2014 ;; MSG SIZE rcvd: 58
Harlan DNS is slower. Over 100ms.
[root@servo rday]# dig finninday.net ; <<>> DiG 9.9.4-P2-RedHat-9.9.4-12.P2.fc20 <<>> finninday.net ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 61312 ;; flags: qr rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 2, ADDITIONAL: 2 ;; QUESTION SECTION: ;finninday.net. IN A ;; ANSWER SECTION: finninday.net. 410 IN A 216.99.216.99 ;; AUTHORITY SECTION: finninday.net. 79345 IN NS yns2.yahoo.com. finninday.net. 79345 IN NS yns1.yahoo.com. ;; ADDITIONAL SECTION: yns1.yahoo.com. 63133 IN A 67.195.1.92 yns2.yahoo.com. 63133 IN A 98.139.247.192 ;; Query time: 117 msec ;; SERVER: 192.168.0.1#53(192.168.0.1) ;; WHEN: Fri May 02 16:09:15 PDT 2014 ;; MSG SIZE rcvd: 126
measure bandwidth with curl
The curl manual says the speed is measured in bytes per second.
Over a wired link, I get this:
root@ferret:~# curl -o /dev/null http://speedtest.wdc01.softlayer.com/downloads/test500.zip % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 20 500M 20 103M 0 0 5683k 0 0:01:30 0:00:18 0:01:12 6896k^C
Over Portlandia, I get about 2MBps out of a theoretical max of 6.75MBps:
[rday@servo lf]$ curl -o /dev/null http://speedtest.wdc01.softlayer.com/downloads/test500.zip % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 7 500M 7 38.2M 0 0 1595k 0 0:05:20 0:00:24 0:04:56 1936k^C
Over ahub, I get about 2.6MBps out of a theoretical max of 4.5 to 6.75MBps:
[rday@servo lf]$ curl -o /dev/null http://speedtest.wdc01.softlayer.com/downloads/test500.zip % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 11 500M 11 56.3M 0 0 2397k 0 0:03:33 0:00:24 0:03:09 2632k^C