Feeder Sometimes Stops at 00:00 for a while

[verdi]

FR24.jpg Anyone have any thoughts as to what is happening and why my feeder does this sometimes ?

[Roeken City]

Hi, I have the same issue. I contacted my internet provider by mail today. So I am waiting for an answer.

[verdi]

FR24.jpg I will be interested in what your provider might have to say. I checked the router logs and don't see any issue. My feeder did it again last evening. The reported availability was down to 70+ % even though it doesn't show that it was offline that long. This morning it is still only showing 98% even though it supposedly has been running uninterrupted all night.

[wiedehopf]

I'd say it's fr24feed auto update .... ;P
But only a guess, no evidence.

I'd tell you to check the logs for the service ... but setting up logs and everything is somewhat complicated, maybe someone else can tell you how to do it.

[Roeken City]

I can't imagine it's FR24 related. 4 days in a row 14 - 18 minutes downtime. I checked my modem and router too and no issues either. Still no reaction from my provider, they are still looking for a good excuse I presume. Screenshot Datafeed.jpg

[Oblivian]

Rule number one on stats page

And I/users have said before...

Don't rely on stats page
Unless it's more than a few hours and you get an email

It's not going to effect business status in that time
Its 'come back'
It's a different set of servers doing work
It's a different upload stream than plane/received data
It's not a realtime udp connection like FA so subject to hiccups

If You think there's a problem with your feed. Don't use a remote website to confirm it. Use local data/logs.

[Roeken City]

This morning my internet provider came with a brilliant excuse. They are too busy due to the pandemic, so they can't help me. I am lucky my contract ends in a few months, so I am going to look for a better one.
In the meantime I analysed my Pi-Hole data trafic during the FR24 "downtime". Result: no issues found.
Unless someone comes with a better idea I think the former speaker used the right words: "Don't rely on stats page".

[Oblivian]

Yep.

If your local router logs show no issue. And the various feeder logs show now issue. It's most likely the way the logs are occasionally updated or accumulated as part of the feeder.

There is also an auto-updater if you did not turn it off. Or auto-restart when no traffic received for a period (an attempt to self-heal) during quite times. Which can restart the feeder very quickly. But accentuate the actual recovery time on the stats page incorrectly.

[peerke63]

Thanks Verdi and Oblivian...

Have same "problem" as first post.
But reading your posts Oblivian, I'll leave it be

I only see 2 spikes in Disk I/O IOPS and Bandwith at midnight.


Graphs (2).jpg

[G6CMG]

Just Started seeing the exact same thing 14-19 Min gaps just after midnight over the last week, with all my other feeders showing no problem and Grapsh1090 showing nothing out of the ordinary during the reported offline gaps!

Running updated Raspberry Buster with all the latest updates, hard wired to router and FR24 S/W Ver 1.0.27-2

[diezel]

I too have outages for the past 4 days from 00:00 to 00:15 approximately.

I believe this to be service related on the FlightRadar24 server side, not so much our client side.

FWIW: I have recently enabled Log Rotation via rsyslog and logrotate, I do not believe these issues to be tied to each other, as I enabled log rotation 2 weeks ago.


Consulting the log file of my feeder: /var/log/fr24feed/fr24feed.log Mine is looking like a restart of the feeder, and subsequent HTTP 520 & HTTP 502 error codes when contacting the FlightRadar24 servers:


Thanks to how terrible the forum software is here, I can't even post snippets of the error logs. Nice.

[Oblivian]

Even the forums are timing out during the day and giving 524 (100ms host timeout) messages or just slow and stops serving pages.

wouldn't be too worried.

[neroon79]

Although it is about 2 weeks after last post I'd like to add the following: I coinidentically checked my stat page during one of these gaps. I went to receiver status an it showed green. I monitored it for about 15minutes until I went to bed. Checking my stats page on next day the gap lastet at least 10 further minutes after I started monitoring feeder status which showed green during the whole time. So problem seems not to be at my end.

Well, if you look on global stats there is no receiver shwing the theoretical 720hour online time forpast 30 days. The best ones stuck at 718 +/-1 hour. Bevore the gaps started to happen many receivers were showing 720 of 720 possible hours (mine too by the way). So if obviously everyone is affected any further investigation at receiver side makes no sense.

Since today morning all stations which have switched on statistic sharing are showing no recent data available for this radar (20 checked). First check at 0200UTC, last check 0500UTC.

[verdi]

My FR 24 feeder supposedly stopped sending data at 00:00 last evening and is still not sending data. However, it says that it is on line. I have not received an email telling me that my feeder has gone down. The FR 24 status page indicates that my receiver is tracking AC. Did a SW restart but that didn't help. Says its online, connected and tracking AC but apparently not communicating with the mother ship.....

[T-NZMB95]

I experienced similar issues until about 1 week ago along with dropouts at other times.
After changing my Pi to Ethernet the dropouts appear to have stopped.
I also experienced dropouts on WiFi connected streaming devices (Amazon Echo and Apple TV) so I changed the WiFi Channel in my Router by 1 Channel (2.4Ghz 11 to 10) and it appears that it was Local WiFi Interference.