SSID Overload

[jonakarl]

Not sure how to fix this but maybe it could be incorporated in the test schedule when the nodes are assembled that they also shut off the ssid broadcast (or update the software) ?

As it is now, every time we power up new nodes we get a semi angry mail from IT that we have X amount of rouge access points clogging up the airways.

[kristrev]

This was discussed in the plenary, telcos and on email. It was agreed to keep wifi on, but with a log signal strength to reduce interference, as it is required for draining the battery in case a mifi crash. Without wifi, it takes a very, very, very long time to drain the battery (as opposed to long).

We can always rename them to Jonas Karlsson's iPhone though ;) We might diable wifi later though.

[jonakarl]

Ok I totally missed that discussion (or forgot it) but does it need the SSID broadcast for draining the battery?

Anyway, this could potentially be a problem with customers that do not allow rouge APs but maybe that could be handled case by case.

Please don't rename them to that, the complaints end up in my email box whatever they are called but, I would be slightly offended by being connected with apple products though :)

[jonakarl]

I have spoken with Tom and it would be very good if you (temporary until the nodes are deployed) disable wifi on the following nodes (ie all new) : 81
87
91
92
93
95
96
98
99
104
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251 So far I have heard no complaints about people no being able to connect or use the wireless network but if the complaints start to drop in we are the main target and I think it is better to prevent that. Having wifi off (and loose the ability to drain the batteries on the mifis) is manageble when the nodes are located in the building and then we can enable the wifi once we deploy them.

[kristrev]

Most of us are busy with other tasks that higher priority this week, so I guess you are on your own when it comes to this. 5GHz can be disabled through the UI, 2.4 GHz requires a script we wrote and which is on Github.

However, I guess they can't really do anything but complain. Spectrum is after all unlicensed :)

[jonakarl]

Ok, we will have a discussion internally but in worst case I will power off the mifis/nodes until you have time to fix this (I cannot remote control the nodes as you do even if I have access to the script). By the way they do can force us to shut down the nodes and never start them again and also ban us from starting new nodes that do not have wifi off by default (we are in the same "company" so they have all the rights).

2016-09-19 14:26 GMT+02:00 Kristian Evensen notifications@github.com:

Most of us are busy with other tasks that higher priority this week, so I guess you are on your own when it comes to this. 5GHz can be disabled through the UI, 2.4 GHz requires a script we wrote and which is on Github.

However, I guess they can't really do anything but complain. Spectrum is after all unlicensed :)

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-247979276, or mute the thread https://github.com/notifications/unsubscribe-auth/AItal7A72HZ9e9_ShNhJk_doF9CGm1D2ks5qrn-IgaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE

[relet]

Hi Jonas -

If you can convince the IT staff that the wifis will not pose an issue (except for showing up with a skull and bones in the interface, and possibly hiding any worse culprits) that would be the best solution. They will always be the lowest powered WIFI in any given situation, and should not interfere with a network in a reasonable distance.

I can run the disable wifi script on all nodes, but that takes time and relies on the mifis actually being up and running. We also lose the option to quickly reboot the mifi, instead of draining the battery for a reset, when USB is unavailable. Then, we will have to revert the effort when deploying the nodes.

Thank you!

[alfs]

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/training/resources/adjacent-channel-congestion.html)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

[realmobile]

The problem with wifi should be gone when you deploy the nodes with hosts?

1. sep. 2016 kl. 08.25 skrev Stefan A notifications@github.com:

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/training/resources/adjacent-channel-congestion.html http://www.metageek.com/training/resources/adjacent-channel-congestion.html)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248213286, or mute the thread https://github.com/notifications/unsubscribe-auth/ADptul5bnbSZj-dgJkMyd6SJf3Dxv8mMks5qr3xOgaJpZM4KAUU-.

[jonakarl]

Maybe maybe not, that depends on the hosts.

I think almost all medium to big company's have (or should at least have) a policy against rouge APs. Either due to that it mess up the airways and might create air time congestion or maybe more likely as it can be considered a security risk (backdoor inside the firewall)

At our University IT are pretty understanding about this but I would not be surprised if at some company a unregistered AP would result in a direct ban of that device.

2016-09-20 8:30 GMT+02:00 realmobile notifications@github.com:

The problem with wifi should be gone when you deploy the nodes with hosts?

1. sep. 2016 kl. 08.25 skrev Stefan A notifications@github.com:

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/ training/resources/adjacent-channel-congestion.html < http://www.metageek.com/training/resources/adjacent- channel-congestion.html>)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213286>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptul5bnbSZj- dgJkMyd6SJf3Dxv8mMks5qr3xOgaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248213994, or mute the thread https://github.com/notifications/unsubscribe-auth/AItal41RUpfIXPC4wpAec4irI2xC_-paks5qr31wgaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE

[realmobile]

So that means that a visitor or and employee using his iphone in Mobile hotspot mode will be kicked out? I do not think that 3 mobile hotspots in any environment will cause any problems. I understand it when we have 60 in our lab, but when it is deployed and limited to 3, I will be very surprised if that causes any problems.

Audun

1. sep. 2016 kl. 08.40 skrev Jonas Karlsson notifications@github.com:

Maybe maybe not, that depends on the hosts.

I think almost all medium to big company's have (or should at least have) a policy against rouge APs. Either due to that it mess up the airways and might create air time congestion or maybe more likely as it can be considered a security risk (backdoor inside the firewall)

At our University IT are pretty understanding about this but I would not be surprised if at some company a unregistered AP would result in a direct ban of that device.

2016-09-20 8:30 GMT+02:00 realmobile notifications@github.com:

The problem with wifi should be gone when you deploy the nodes with hosts?

1. sep. 2016 kl. 08.25 skrev Stefan A notifications@github.com:

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/ training/resources/adjacent-channel-congestion.html < http://www.metageek.com/training/resources/adjacent- channel-congestion.html>)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213286>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptul5bnbSZj- dgJkMyd6SJf3Dxv8mMks5qr3xOgaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248213994, or mute the thread https://github.com/notifications/unsubscribe-auth/AItal41RUpfIXPC4wpAec4irI2xC_-paks5qr31wgaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE — You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248216005, or mute the thread https://github.com/notifications/unsubscribe-auth/ADptum3juNkMbH9amxeU2r0v1NimK9wWks5qr3_3gaJpZM4KAUU-.

[jonakarl]

It of course depends on the company but yes some companys have policy's against phones in "hotspot mode" or unregistered APs in general and will result in a "your boss needs to talk with you". However, if the device is not connected to the internal network and with only 3 mifis at low power, I also do not think it is a technical problem but policy wise it might still be a problem, lets wait and see.

All this said I do not think we need to take any action on preventing this once the nodes are deployed but more to be aware that it might be a problem and have routines for handling this (ie have simple and quick procedures to turn off wifi on nodes where it is requested).

However, as you said when we have 30*3 APs in our lab it is a problem.

I also think that Stefans suggestion to lock the channel to say 1 on all the modem would help in mitigating the technical problem that might a rise as we then at least leave 3 and 11 as free channels to use.

2016-09-20 8:44 GMT+02:00 realmobile notifications@github.com:

So that means that a visitor or and employee using his iphone in Mobile hotspot mode will be kicked out? I do not think that 3 mobile hotspots in any environment will cause any problems. I understand it when we have 60 in our lab, but when it is deployed and limited to 3, I will be very surprised if that causes any problems.

Audun

1. sep. 2016 kl. 08.40 skrev Jonas Karlsson notifications@github.com:

Maybe maybe not, that depends on the hosts.

I think almost all medium to big company's have (or should at least have) a policy against rouge APs. Either due to that it mess up the airways and might create air time congestion or maybe more likely as it can be considered a security risk (backdoor inside the firewall)

At our University IT are pretty understanding about this but I would not be surprised if at some company a unregistered AP would result in a direct ban of that device.

2016-09-20 8:30 GMT+02:00 realmobile notifications@github.com:

The problem with wifi should be gone when you deploy the nodes with hosts?

1. sep. 2016 kl. 08.25 skrev Stefan A notifications@github.com:

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/ training/resources/adjacent-channel-congestion.html < http://www.metageek.com/training/resources/adjacent- channel-congestion.html>)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213286>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptul5bnbSZj- dgJkMyd6SJf3Dxv8mMks5qr3xOgaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213994, or mute the thread https://github.com/notifications/unsubscribe-auth/ AItal41RUpfIXPC4wpAec4irI2xC_-paks5qr31wgaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE — You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248216005>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptum3juNkMbH9amxeU2r0v1NimK9 wWks5qr3_3gaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248216605, or mute the thread https://github.com/notifications/unsubscribe-auth/AItal1loidDLTseztur0KDukzhZpM8Q9ks5qr4C6gaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE

[ozgualay]

When are you planning to distribute the nodes, I think we should start the process at least for the stable ones. This will also help solve the issue one way or another :)

Özgü

On 20 September 2016 at 09:09, Jonas Karlsson notifications@github.com wrote:

It of course depends on the company but yes some companys have policy's against phones in "hotspot mode" or unregistered APs in general and will result in a "your boss needs to talk with you". However, if the device is not connected to the internal network and with only 3 mifis at low power, I also do not think it is a technical problem but policy wise it might still be a problem, lets wait and see.

All this said I do not think we need to take any action on preventing this once the nodes are deployed but more to be aware that it might be a problem and have routines for handling this (ie have simple and quick procedures to turn off wifi on nodes where it is requested).

However, as you said when we have 30*3 APs in our lab it is a problem.

I also think that Stefans suggestion to lock the channel to say 1 on all the modem would help in mitigating the technical problem that might a rise as we then at least leave 3 and 11 as free channels to use.

2016-09-20 8:44 GMT+02:00 realmobile notifications@github.com:

So that means that a visitor or and employee using his iphone in Mobile hotspot mode will be kicked out? I do not think that 3 mobile hotspots in any environment will cause any problems. I understand it when we have 60 in our lab, but when it is deployed and limited to 3, I will be very surprised if that causes any problems.

Audun

1. sep. 2016 kl. 08.40 skrev Jonas Karlsson <notifications@github.com :

Maybe maybe not, that depends on the hosts.

I think almost all medium to big company's have (or should at least have) a policy against rouge APs. Either due to that it mess up the airways and might create air time congestion or maybe more likely as it can be considered a security risk (backdoor inside the firewall)

At our University IT are pretty understanding about this but I would not be surprised if at some company a unregistered AP would result in a direct ban of that device.

2016-09-20 8:30 GMT+02:00 realmobile notifications@github.com:

The problem with wifi should be gone when you deploy the nodes with hosts?

1. sep. 2016 kl. 08.25 skrev Stefan A notifications@github.com:

But IT staff think it does pose issues, quote (swedish, sorry)

Även om SSID:et är dolt så skickas det ut beacons om att det finns, varje beacon belastar radioutrymmet, nu är det iofs i 2.4 Ghz som redan är ett frekvensband som är helt förstört, vi ser att fler och fler klienter ansluter till 5Ghz.

Sedan ser jag att de använder kanal 3,4,5,6,7,8,9 och 10, detta ger oss Adjacent channel congestion (http://www.metageek.com/ training/resources/adjacent-channel-congestion.html < http://www.metageek.com/training/resources/adjacent- channel-congestion.html>)

De access punkter vi har i 21E korridoren ser att vissa av monroe-noderna har ganska hög effekt -46dBm."

Would it be possible to lock to a single channel (e.g. 1) instead of autoselecting? Are we sure that they transmit at the lowest power? (measurement above indicated otherwise)

Could wifi on/off be automated, e.g. a configurable option in the inventory, that the node would check at reboots and act on when modems are available?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213286>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptul5bnbSZj- dgJkMyd6SJf3Dxv8mMks5qr3xOgaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248213994, or mute the thread https://github.com/notifications/unsubscribe-auth/ AItal41RUpfIXPC4wpAec4irI2xC_-paks5qr31wgaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE — You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248216005>, or mute the thread https://github.com/ notifications/unsubscribe-auth/ADptum3juNkMbH9amxeU2r0v1NimK9 wWks5qr3_3gaJpZM4KAUU-.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17# issuecomment-248216605, or mute the thread https://github.com/notifications/unsubscribe-auth/ AItal1loidDLTseztur0KDukzhZpM8Q9ks5qr4C6gaJpZM4KAUU- .

Jonas Karlsson Senior Research Engineer

Karlstad University SE-651 88 Karlstad, Sweden Telephone: +46 54 700 15 64 Mobile: +46 70 672 06 20 Skype: karlsson.karl.jonas Hangout: karlsson.karl.jonas

twitter.com/kau facebook.com/karlstadsuniversitet KAU.SE

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248221350, or mute the thread https://github.com/notifications/unsubscribe-auth/AB_G4Lq_F_hVGyx4jCC9Jrkkr8Ss6fteks5qr4a7gaJpZM4KAUU- .

[relet]

Just to confirm, this is from what one of the nodes picked at random (#242/usb0) self-reports (mf910-info utility): "wifi_coverage": "short_mode", "m_ssid_enable": "0", "ap_station_mode": "wifi_pref" "SSID1": "864154026014845",

not sure what these fields are, just for the sake of completeness: "m_SSID": "4G-Gateway-5E6D_2", "m_HideSSID": "0", "EX_SSID1": "0001softbank",

[kristrev]

Is the internal wifi switched on on the node? Since the internal wifi is also configured to act as an AP with hidden SSID, I suspect it could be the one they complain about (strong signal).

[relet]

No, the internal mifi access point is not enabled by default in the configuration.

On Wed, 21 Sep 2016, 09:57 Kristian Evensen, notifications@github.com wrote:

Is the internal wifi switched on on the node? Since the internal wifi is also configured to act as an AP with hidden SSID, I suspect it could be the one they complain about (strong signal).

— You are receiving this because you commented.

Reply to this email directly, view it on GitHub https://github.com/MONROE-PROJECT/Maintenance/issues/17#issuecomment-248538902, or mute the thread https://github.com/notifications/unsubscribe-auth/AAISEAJAdePqu5q70qY3CIBnnU622c34ks5qsON0gaJpZM4KAUU- .

[jonakarl]

Not relevant after retiring the mifis and big "node rebuild".