Announcement

Collapse
No announcement yet.

very high load rpi

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • very high load rpi

    PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
    474 fr24 20 0 130540 8836 4668 S 293.8 0.9 323:53.26 fr24feed
    2533 fr24 20 0 25872 7348 2136 R 18.8 0.8 34:04.93 dump1090-mutabi


    This happens once a week, there are only 7 aircraft seen so thats not an issue, and its been ongoing since slightly after 0400 local - when curfew is active so at best it might have seen 1 a/c if any. sometimes it kills off dump1090

    there are no crons anywhere near that time, and i dont log since that too created problems when it tries to roll the 2g file, it never seemed to rotate every day or 2 days.

    yes, apt is up to date.

  • #2
    2GB! Yowzers!

    There's a logrotate script in this thread that works well.

    Comment


    • #3
      nice feature fr24 should add that to their install page, and yes I get the logfile rotate is broke - hence why I disabled it.
      But the logfile is not the cause of this once a week sky rocketing load, I guess it has something todo with that evil shit that is systemd that has its own time control that fr24 have told it to do some crap at some time

      I will one day try get slackware to run on the pi, as i have no systemd virus infected machines here until hte pi and would prefer to keep it that way.

      either way though, there must be something fr24feed is doing weekly that causes this.

      Comment


      • #4
        OK so it seems this happens around 0030 localtime every 6 or so days.
        dump12090-mutability is not running.
        I note if i kill it off manually, fr24feed detects this and restarts it
        but in these cases, it is not restarting...

        fr24 want to comment on what the pi image of yours does at around that time?
        There is no drop outs of links, either ethernet, or upstream which link has been up for 427 odd days.

        Something in your image is causing this, killing it off and and is unable to restart dump1090, seemingly looping trying to ending up with high loads, high temps and so on...

        I dont do any logging now for couple weeks so rotating log files cant be an issue as I thought it might have been, there is no cron jobs at all for fr24feed.

        rather annoying ongoing issue.

        I have put in place a temp fix, but would like the root cause to be identified and resolved as this is not really ideal but it corrects the issue

        #!/usr/bin/perl
        use strict;
        use Sys::Syslog;
        use Mail::Sendmail;
        my $fr24 = `pgrep fr24feed`;
        my $d1090run = `pgrep dump1090`;

        if (! $d1090run) {sleep(30);}

        if (($fr24) && (! $d1090run)) {
        print "dump1090 is not running, trying to restart\n";
        syslog('info',"dump1090 is not running, trying to restart");
        if (system("sudo systemctl restart fr24feed") == 0) {
        syslog('info',"Success");
        print "Success\n";
        } else {
        syslog('info',"Still in error state");
        print "Still in error state\n";
        &mailme;
        }

        }

        Comment

        Working...
        X