FlightAware Discussions

Announcing PiAware 5!

Today we are pleased to announce the release of PiAware 5!

Things to note:
· 978 installations are strongly encouraged to upgrade; this version fixes a bug affecting all older versions that produces garbage messages
· 1090 installations should see a slightly increased message/position rate
· 1090 installations will see increased CPU; this is an expected consequence of the improvements made above
· New SkyAware URL: /skyaware or :8080 . If you only have dump978 installed, you will continue to use /skyaware978 which has the latest features
· The existing /dump1090-fa/ and /skyaware978/ maps and URLs are still available, although will be deprecated in a later release as we work towards one SkyAware package containing all the front end interface
· PiAware 5 will be the last version to support Raspbian Jessie OS. Upstream support has stopped and no longer has security fixes, updates, etc.

· If you are using upintheair.json feature, you will need to add your json to /usr/share/skyaware/html/ to get this to work with the new merged map

How to Upgrade

  • Existing SD card installs: Select “Upgrade and Restart PiAware” via your My ADS-B stats page Site Configuration menu (gray gear icon)
  • Existing Package installs: Select Upgrade PiAware AND Upgrade dump1090-fa and/or dump978-fa via your My ADS-B stats page Site Configuration menu (gray gear icon)
  • PiAware 5 SD card download is here.
  • PiAware 5 LCD SD Card download is here


What’s New?

SkyAware
· New Altitude, Speed, Aircraft Ident, and Aircraft Type filters to fine tune the select aircraft you want see
· Ability to toggle visibility of ADS-B, MLAT, UAT, TIS-B, and Other aircraft on SkyAware using the check boxes under the aircraft table
· Ability to show Aircraft Ident labels for all aircraft on your display (found in the View Toggles list in the settings menu)
· A merged SkyAware map of 1090 and UAT aircraft for those with dual 1090/978 setups.
· Direct link on the SkyAware map to go to your My ADS-B statistics page on FlightAware.com or a direct link to claim your receiver if it has not been claimed (SD card installs only)
· Bold aircraft Idents are used to indicate tail numbers in the aircraft table rather than using the previous prepended underscore syntax

dump1090-fa
dump1090: limesdr: use --gain value if --limesdr-gain is not specified
dump1090: Switch sample-conversion DSP code to use starch. Provide NEON implementations for ARM when the hardware supports it. Update 8-bit unsigned zero offset to reflect actual rtlsdr behavior. This gives a substantial speedup to sample processing for the 12/16-bit case, and a small speedup for the 8-bit case. Default wisdom provided for arm, aarch64, and x86; see the README for details on generating hardware- specific wisdom files.
dump1090: Make console stats display fixed-width (courtesy @gtjoseph)
dump1090: Add --json-stats-every option to control the update cycle of stats.json
dump1090: Don’t expire data when reading a capture at full speed via --ifile
dump1090: Count received messages for each DF type, emit in console stats / stats.json
dump1090: Write final stats.json on process exit (mostly useful for --ifile)
dump1090: Attempt to correct DF11/17/18 messages where the DF field itself is damaged. This requires a lot more CRC checks when receiving messages off-the-air, but slightly increases the decode rate for DF11/17/18.
dump1090: Add a --no-fix-df option to disable the extra DF field correction, for cases where the extra messages are not worth the extra CPU
dump1090: Track recently-seen DF18 addresses, but only count them as recently-seen for DF18 decoding, not for Mode S / DF17

dump1090: Support --stats-every intervals of less than 1 minute correctly
dump1090: Don’t erroneously consider ICAO 000000 as “recently seen” when decoding
dump1090: Fix rare hang (1 in 1e9) in FIFO code due to an off-by-one error in normalize_timespec
dump1090: Correctly expire stale emergency, nac_v data (courtesy @wiedehopf, issue #96)
dump1090: Fix FIFO bounded memory leak / effective reduction in size of the FIFO, triggered if the demodulator couldn’t keep up with the input rate
dump1090: Set at least airground = AG_UNCERTAIN for airborne position messages with valid altitudes (issue #113)

view1090
view1090: Interactive mode enhancements to show min/max RSSI, distances (courtesy @gtjoseph, PR #81)
view1090: Fix table alignment with --interactive-ttl > 99 (courtesy @MavEtJu, PR #99)
view1090: Tweak Mode column to display more data-source info
view1090: Fix librtlsdr output interfering with interactive mode display (courtesy @VasiliyTurchenko, PR #116)

general:
Fix linking against NetBSD curses library (courtesy @jvanwouw, PR #111)

dump978-fa
dump978: Fix bugs in the FEC library that could incorrectly mark uncorrectable messages as successfully corrected.
dump978: Add a new network message format for metadata-only messages. This format should be silently ignored by legacy clients; it can be disabled entirely by using the --raw-legacy-port option
dump978: On new network connections, emit a metadata message containing decoder name and version, and a flag indicating the FEC fixes are present
faup978: Report software name/version/FEC fix flag upstream when known

5 Likes

Just a reminder for those with piaware package installs, you’ll need to update dump1090-fa (and dump978-fa if applicable) via the stats page as well to pull in the new dump1090-fa and SkyAware changes.

Let us know if you find any issues!

4 Likes

I’ve updated dump-1090 and Piaware on two RPis and no longer have Skyaware (+other programs)

I have an early Raspberry Pi Zero and although it is at version 4.0 I get this: [Site 56684]

Release file for http://archive.raspberrypi.org/debian/dists/buster/InRelease is not valid yet (invalid for another 40min 42s). Updates for this repository will not be applied.

This usually happens when the system time is off on the Pi. Can you see what date returns?

Upgraded my 2 stations, one with RTL-SDR dongle went fine, but the one with the Airspy mini not working at all.
On the Airspy mini, no maps, nothing, as though the web server is not running. On virtual Radar I don’t see it either, yet the log on FlightAware seems to show it is receiving messages… guess I will have to dig a bit.

It is an hour and a half slow. How do you correct? Edit: Sorted by installing NTP.

Successfully upgraded to ver 5.0 by RE-IMAGING

(1) Stn 5252 - RPi Model 2 / Piaware SD card image v 5.0

1.1 - Down loaded and wrote following image to microSD card
http://piaware.flightcdn.com/piaware-sd-card-5.0.img.zip

1.2 - Followed instructions here
How to Install and Configure Piaware 5.0 SD card image - Quickstart Guide

 

(2) Stn 6396 - OrangePiPC / Armbian Buster

1.1 - Down loaded and wrote following image to microSD card
https://redirect.armbian.com/orangepipc/Buster_current

1.2 - Followed instructions here
[Package Install] How to Install and Configure Piaware on Raspberry Pi OS Image / RPi

 

(3) Stn 76000 - RPi Model 4 / 64-bit Raspberry Pi OS

1.1 - Down loaded and wrote following image to microSD card
https://downloads.raspberrypi.org/raspios_arm64/images/raspios_arm64-2020-08-24/2020-08-20-raspios-buster-arm64.zip

1.2 - Build & Installed piaware and dump1090-fa using following bash scripts

PIAWARE

sudo bash -c "$(wget -O - https://raw.githubusercontent.com/abcd567a/piaware-64bit-raspbian/main/build-and-install-piaware.sh)" 

NOTE: Please be patient. This will take very long time, as the script first installs build tools & dependency packages, and after that downloads source code & builds piaware from source code.

 

DUMP1090-FA

sudo bash -c "$(wget -O - https://raw.githubusercontent.com/abcd567a/piaware-64bit-raspbian/main/build-and-install-dump1090.sh)"  

NOTE: Please be patient. This will take very long time, as the script first installs build tools & dependency packages, and after that downloads source code & builds dump1090-fa from source code.

 

dump1090-fa

 

dump978-fa

2 Likes

I have to say that lighttpd is the most annoying piece of software. Every time piaware updates the config breaks for one reason or another and I have to rummage around to find out what has broken. In this case there was a duplicated server.modules += ( "mod_setenv" ) and a duplicated \data section that was preventing it start. It’s probably OK if there is only one thing using it, but if you have any other software installed it then it’s highly likely to be problematic.

Edit - just noticed the other thread talking about this issue if people look have the problem and look here first.

2 Likes

Early days yet but I am seeing a 15% increase in CPU? Others see the same?

Overall CPU as below:

1 Like

To avoid all this hasstle, I simply re-image microSD card. No issues :slight_smile:

Re-imaging Made Easy

 

Screen Shot 2021-03-12 at 08.59.37

Apologies in advance for the pedantic niggle…

Do you think you could set the PiAware status button to orange for those of us still running 4.0?

To me red implies PiAware has fatal issues, and the older version is still running quite happily. In this case it is just an advisory, not requiring immediate action.

I’m not planning to upgrade for a few days until the main issues are fixed or worked around, and seeing a red status flag every time I look at my stats page is annoying.

7 Likes

That is significantly more hassle than just fixing lighttpd given the location of the pi and what else is installed on it.

2 Likes

IIRC you are running an AirSpy with Progs decoder.
Did CPU usage increase much with the update to PiAware 5?

I’m running an Airspy mini on an Odroid XU4 and not noticed any increase in CPU usage after the update.

image

i have an SD card install and running AirSpy mini. with PiAware 5 i see maybe a 2% to 3% increase in CPU usage…barely noticable

This is true if Pi is not physically accessable easily, and/or one has installed speciality hardware/software like AirSpy dongle etc.

In my case all my Pi’s are easily accessable physically, and I dont have speciality hardware/software. No AirSpy dongle, only FA Pro Plus and RB Flight Sticks. For me reimaging was the easiest and fastest solution.

I also had many problems was using 4.1 but after 3 hours all fixed thanks ro rhe forums here
many thanks to all
pi x 2 version 5

I’m seeing an 8%ish increase in cpu usage.