New monitoring script in 2.1.000 release

branko · 2015-12-15T19:46:42.994Z

The monitoring script has been completely changed and released in the 2.1.000 hotfix. I’ll explain here what the changes are and why we had to change it.

First of all, the reason for the change. We have decided to reduce the payload size in order to support low-bandwidth or low-data cap connections in remote areas. A single data point is now around 34 bytes as opposed to around 230 bytes in the previous version. Unlike the previous version, the data is transmitted in binary (although it is still stored in readable JSON format).

The new data format consist of the following fields:

start marker (OHD, short for Outernet Heartbeat Datagram)
device ID (128-bit identifier reasonably unique to each receiver)
timestamp (4 bits, low-resolution creation time relative to previous heartbeat datagram)
tuner vendor ID (16-bit)
tuner model ID (16-bit)
tuner preset ID (5 bits, finger preset of the tuner settings used to identify the satellite in use)
signal lock (1 bit)
service lock (1 bit)
signal strenght (4 bits)
SNR/quality (5 bits)
bitrate (6 bits)
carousel count (5 bits)
carousel status (31 bits for tracking up to 31 carousels, on/off for each)
end marker (DHO, reverse of the start marker)

The source code can be found on GitHub if you have concerns about what we collect and what we do with it. The monitoring script is in readable format on your receiver.

NOTE: Monitoring can be disabled by removing the file /mnt/persist/monitoring.yes. To do that, SSH into your device and run the following command:

sudo rm /mnt/persist/monitoring.yes

There is no need to restart your device. Monitoring will stop collecting and transmitting data as soon as you remove the file.

kalashnikov · 2015-12-15T23:22:26.396Z

and new status web page http://status.outernet.is/

Syed · 2015-12-16T00:31:48.119Z

When I perform a USB update to 2.1, after I associate with the network, I see a 500-error. When I reboot, I no longer see this problem.

kenbarbi · 2015-12-16T13:31:55.117Z

I saw the 500 error too. I think I didn’t wait long enough after the first power cycle of the Lighthouse. I gave it 5 minutes getting a coffee refill, and it came up with the setup screen.

Now I’ll wait to see what magical data comes down from Galaxy 19

Ken

branko · 2015-12-16T14:09:29.056Z

Yeah, I never got around to setting up a more informative page that would possibly automatically refresh when Librarian is ready…

branko · 2015-12-16T15:53:43.313Z

A post was split to a new topic: Comments on new Outernet content selection

kenbarbi · 2015-12-16T16:55:39.954Z

Branko, this where I found the 2.0.001 Hot Fix We Moved! going there from your announcement page. Ken

kenbarbi · 2015-12-16T17:13:40.974Z

Oh I see the confusion - - I called it 2.0.001 with Hot Fix. I did download the correct versions named 2.1.000, and load it on my Lighthouse.

For the Lighthouse I got it at: We Moved!

For ORxPi (not tested yet) I got it at:
https://archive.outernet.is/images/ORxPi/2.1.000/

Lighthouse shows the new version now:

It is interesting what Pradeeka said about his Tweets - - he is also getting the Ghost Tweets so it’s not a satellite download problem. Ken

pradeeka · 2015-12-16T17:21:39.929Z

kenbarbi:

Oh I see the confusion - - I called it 2.0.001 with Hot Fix.

yes. this is little bit confused. see the position of ‘1’ carefully.This is a common problem with our EYEs.

branko · 2015-12-16T17:24:36.864Z

The reason for that weird and ugly version number is a limitation in the sort utility in the OS which doesn’t recognize version numbers. Ideally, we’d use something like 2.1a1 or 2.1RC1 or something more human-friendly, but the sort would treat some of those as newer than 2.1, and not allow update from alpha/RC versions. So we dcided to have a 3-part version number where the last 3 digits are used for development snapshots.

pradeeka · 2015-12-16T17:31:42.232Z

My LH is now up and running with 2.1.000 hot fix. But the status is not displaying on the status page. It should be ABS-2.

Pasted image1366×728 50.5 KB

branko · 2015-12-16T17:33:09.655Z

pradeeka:

My LH is now up and running with 2.1.000 hot fix. But the status is not displaying on the status page. It should be ABS-2.

We still haven’t plugged that issue, but did not want to delay the hotfix until that was fixed. We are currently working on a new algorithm for detecting reception problems as we’ve gained significantly more experience since the first monitoring server was implemented.

pradeeka · 2015-12-16T17:36:49.479Z

then how Galaxy 19 and Hotbird 13 status are up?

branko · 2015-12-16T17:38:37.722Z

They sometimes turn off.

kenbarbi · 2015-12-16T17:56:01.121Z

Would I be correct in assuming, since I am on Galaxy 19, and connected to the internet, your server would carry a report from me, or is the one receiver shown someone else? Seems as if there should be more than just 1 receiver reporting in on Galaxy 19. Ken

Syed · 2015-12-16T19:55:21.268Z

This is odd. Well, it’s a bug. We have a Lighthouse at the office pointed at G19. And there is another one in Boston that we use as a monitor. Not to mention one at my parents’ house. Those should all be online, but are not present in status.outernet

kenbarbi · 2015-12-16T20:48:31.183Z

I ran Branko’s comment:

Pasted image725×413 9.8 KB

and Monitoring is running on my Lighthouse v2.0.000. I don’t know if the output was actually sent out over the internet to anybody. I’ll see if I can determine that. Ken

kalashnikov · 2015-12-16T20:50:27.220Z

hi

LH 2.1.000 on 113w and internet conection…
the status page do not show 113w users …

kenbarbi · 2015-12-16T21:04:31.488Z

Did you try Branko’s check of Monitoring?

grep monito /var/log/syslog

to see if the Monitor program is running? You are using 2.1.000, so it should be running. As I said to Syed, I don’t know if the output is going anywhere. The experts at Outernet will have to figure this one out. Ken

Syed · 2015-12-16T21:37:07.661Z

@kalashnikov What is your bitrate? Did you receive a 500-error the first time you connected?