What have I learnt..in the last few days
I had planned on completing one of these blog posts each day, but it became apparent on Friday that this was clearly a step too far.
Work has been crazy busy at the moment and when I get home I would rather spend time with the family than make blogging my number 1 priority!
Anywho, the plan is to make these a little less regular, maybe once a week.
So what have I learnt this week?
Amazon Dash
Amazon released the Amazon Dash button in the UK last week. The idea is that you put a branded button in certain locations around your home. When you run out of a specific item (e.g. washing powder) you press the Ariel Amazon Dash button on your washing machine. This will contact Amazon and get your Ariel product delivered the next day.
Neat idea, slightly overkill I think, but neat!
But how does it work? How long do they last?
Well I came across this pretty decent teardown blog post from Matthew Petroff.
https://mpetroff.net/2015/05/amazon-dash-button-teardown/
The button only uses a single AAA battery which is soldered in. The way the button works means it only uses minimal energy. The button should easily last longer than the button will be useful.
Interestingly, if you pair the button with an Android device, it will use WiFi. But if you have an iOS device, it uses Ultrasound. That is cool.
Single Cell WiFi
A large part of last week was spent troubleshooting WiFi issues in one of main offices. We use a system from a company called Meru (recently purchased by Fortinet). Their offering utilises a Single Cell architecture.
This means that all Access Points broadcast their SSIDs on the same channel. This can be really useful to easily provision sites as there is less need to do site surveys and plan your channel layout when provisioning a new office. In my experience, office moves usually involve IT being left way down the list people with advance notice, so having something which can be rolled out relatively quickly and require less physical resource is a thumbs up.
But, there is some concerns around throughput which this blog post goes into.
https://community.spiceworks.com/topic/989074-why-zero-hand-off-aka-single-cell-architecture-does-not-work-very-well
Our issue wasn't throughput related, but congestion related and the sheer amount of people in our tiny office. We are running a programme of work to implement an ERP system which involves many people from the business, system integrators and third parties.
In the last week, the WiFi performance has gone off a cliff. Lots of dropped packets between laptops and APs. We used spectrum analyser to see if there was any external interference, but nothing.
We then took a new AP and published a different SSID on a different 5GHZ channel. This worked perfectly. If I moved this SSID onto the 5GHZ channel used by our other APs, the performance was also terrible. This also tallies up with an increase of people who are working in the office on a permanent basis.
So it seems that if you have Single Cell Architecture, be really careful of the density of devices in small offices. Someone suggested more APs, but I think this would just contribute to the problem more than anything.
Our current plan is to change half of the APs to one channel and the other half to something else to try and ease the congestion in the air.
To try and help troubleshoot this, I used the following Powershell script. This would ping Google's DNS servers every 2 seconds and if the response time is greater than 200ms (or timed out) it would add the response time, signal strength, the station connected to and SSID to a csv file. If 5 of these occur in a row, it would add something to the CSV to this affect.
And Finally
HIIT (High Intesity Interval Training) has been discussed as a shortcut to exercise for years. Last week, the BBC ran the following article which is also part of the TV series "Trust Me I'm a Doctor".
http://www.bbc.co.uk/news/magazine-37249021
They ran tests to help prove or disprove HIIT as a credible alternative to the standard mantra of 150 minutes of moderate exercise a week.
The Results were pretty impressive. The group who completed 15 minutes of HIIT across three sessions a week actually had a 17% greater VO2 max than the group who did 150 minutes of moderate exercise.
Pretty impressive, I might make my next run based on Tabata
Sam
Work has been crazy busy at the moment and when I get home I would rather spend time with the family than make blogging my number 1 priority!
Anywho, the plan is to make these a little less regular, maybe once a week.
So what have I learnt this week?
Amazon Dash
 |
| Ariel Dash Button |
Neat idea, slightly overkill I think, but neat!
But how does it work? How long do they last?
Well I came across this pretty decent teardown blog post from Matthew Petroff.
https://mpetroff.net/2015/05/amazon-dash-button-teardown/
The button only uses a single AAA battery which is soldered in. The way the button works means it only uses minimal energy. The button should easily last longer than the button will be useful.
Interestingly, if you pair the button with an Android device, it will use WiFi. But if you have an iOS device, it uses Ultrasound. That is cool.
Single Cell WiFi
A large part of last week was spent troubleshooting WiFi issues in one of main offices. We use a system from a company called Meru (recently purchased by Fortinet). Their offering utilises a Single Cell architecture.
This means that all Access Points broadcast their SSIDs on the same channel. This can be really useful to easily provision sites as there is less need to do site surveys and plan your channel layout when provisioning a new office. In my experience, office moves usually involve IT being left way down the list people with advance notice, so having something which can be rolled out relatively quickly and require less physical resource is a thumbs up.
But, there is some concerns around throughput which this blog post goes into.
https://community.spiceworks.com/topic/989074-why-zero-hand-off-aka-single-cell-architecture-does-not-work-very-well
Our issue wasn't throughput related, but congestion related and the sheer amount of people in our tiny office. We are running a programme of work to implement an ERP system which involves many people from the business, system integrators and third parties.
In the last week, the WiFi performance has gone off a cliff. Lots of dropped packets between laptops and APs. We used spectrum analyser to see if there was any external interference, but nothing.
We then took a new AP and published a different SSID on a different 5GHZ channel. This worked perfectly. If I moved this SSID onto the 5GHZ channel used by our other APs, the performance was also terrible. This also tallies up with an increase of people who are working in the office on a permanent basis.
So it seems that if you have Single Cell Architecture, be really careful of the density of devices in small offices. Someone suggested more APs, but I think this would just contribute to the problem more than anything.
Our current plan is to change half of the APs to one channel and the other half to something else to try and ease the congestion in the air.
To try and help troubleshoot this, I used the following Powershell script. This would ping Google's DNS servers every 2 seconds and if the response time is greater than 200ms (or timed out) it would add the response time, signal strength, the station connected to and SSID to a csv file. If 5 of these occur in a row, it would add something to the CSV to this affect.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 | $errornumber = 0 $outputreport = "c:\temp\reportgoogle.csv" While($a = 1) { Try { Sleep 2 $time = Get-Date -Format T $ping = test-connection 8.8.8.8 -count 1 -ErrorAction Stop | foreach { $_.responsetime} $signal = (netsh wlan show interfaces) -Match '^\s+Signal' -Replace '^\s+Signal\s+:\s+','' $BSSID = (netsh wlan show interfaces) -Match '^\s+BSSID' -Replace '^\s+BSSID\s+:\s+','' $Profile = (netsh wlan show interfaces) -Match '^\s+Profile' -Replace '^\s+Profile\s+:\s+','' Write-Host "Response Time is $ping" $output = $time + "," +$ping + "," + $Signal + "," + $BSSID + "," + $Profile If ($ping -gt 200) { Out-File -FilePath $outputreport -InputObject $output -append $errornumber +=1 } Else{$errornumber = 0} } Catch { $ping = "Request Timed Out" $errornumber +=1 write-host $ping $signal = (netsh wlan show interfaces) -Match '^\s+Signal' -Replace '^\s+Signal\s+:\s+','' $BSSID = (netsh wlan show interfaces) -Match '^\s+BSSID' -Replace '^\s+BSSID\s+:\s+','' $Profile = (netsh wlan show interfaces) -Match '^\s+Profile' -Replace '^\s+Profile\s+:\s+','' $output = $time + "," +$ping + "," + $Signal + "," + $BSSID + "," + $Profile Out-File -FilePath $outputreport -InputObject $output -append Continue } Finally{ $output = "SLOW OVER LAST TEN SECONDS" If ($errornumber -gt 5){Out-File -FilePath $outputreport -InputObject $output -append} } } |
And Finally
HIIT (High Intesity Interval Training) has been discussed as a shortcut to exercise for years. Last week, the BBC ran the following article which is also part of the TV series "Trust Me I'm a Doctor".
http://www.bbc.co.uk/news/magazine-37249021
They ran tests to help prove or disprove HIIT as a credible alternative to the standard mantra of 150 minutes of moderate exercise a week.
The Results were pretty impressive. The group who completed 15 minutes of HIIT across three sessions a week actually had a 17% greater VO2 max than the group who did 150 minutes of moderate exercise.
Pretty impressive, I might make my next run based on Tabata
Sam
Comments
Post a Comment