/var/log/apache2/error.log filling up

2 ASL machines have recently started filling error.log with

gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory
gzip: /tmp/irlpdata.txt.gz: No such file or directory

i dont need anything IRLP related
as I said this is a newish behaviour (maybe 1 month)
maybe a Jan 1 thing?
so how stop it?

de k9wkj

You need to find the source of what is running to know how to stop it.
Look through other log files.
But it looks like an error from some kind of backup program running.
You did not post the time interval of the error. But you might also check your cron or scheduler for something executing on/near the interval.

But likely from something you may have installed and perhaps uninstalled not fully removed.

But if you fail to find it, you might try to satisfy it by creating the file in the said path.

Normally there is nothing related to IRLP on an ASL node. Maybe your some other AllStar. Where did you get your node?

these are both ASL 2.0B6 machines
one is a Pi3b+ on a Quantar the other a Pi4 on a 6M MastrII at different sites on our WAN
these have been running for over a year without apparent log filling issues
I truncated the file and did some other log cleaning earlier in the month
because I had noticed that the node names showed “node not in database” on Supermon, which I tracked to the temp FS being 100% full
but 2 days ago the same thing happened, this time I actually looked at the files contents
I truncated the file again and have checked in on it a few times and now not seeing the behavior
just odd to me that it is a Apache error log
is there some polling of IRLP status being done by Apache ? I snooped a bit but didnt see anything

de k9wkj

I don’t know. Like I said, here’s no attempt to do anything related to IRLP on ASL nodes.

Could it be something with Supermon? It’s likely doing lots of Apache things.

yes this is Supermon as shipped with the image,
except for some visual customizing
I am running a tail on the file on both machines to see if it crops up again
why it started and stopped on its own I dont know
the irlpdata.txt.gz file doesnt exist on the machines

seems like something Supermon/Apache was looking for wasnt there and tossed the error
as I said I snooped a bit in some of the php stuff and didnt see anything
the only IRLP mention I found was in /var/www/html/supermon/allmon.ini
and that is just a link, which I dont have displayed

de k9wkj

so with no changes other than cleaning out the logs, not even a reboot
they stopped exhibiting the behavior
tailed them for a day and see nothing

lets call it closed for now

de k9wkj

If you still have Supermon problems you could try the new and improved Allmon3.

See Allmon3 v1.2.0 Released

just to add to this mystery

I did a run of that really nifty ASL install script that is up on Github
on a Rpi5

Raspberry Pi OS Lite

  • Release date: March 15th 2024
  • System: 64-bit
  • Kernel version: 6.6
  • Debian version: 12 (bookworm)

all went swimmingly
it is very nice to have Dadhi install and run on a current kernel !!!
I let the script install Supermon,Allmon3 and Allscan
all is well except that
gzip: /tmp/irlpdata.txt.gz: No such file or directory
is filling up /var/www/apache2/error.log
with a entry every second
using “iotop” I see it is Apache doing the log filling
so I ran Allmon3,Allscan,Supermon all at once
and determined that it is Supermon behaving this way, just by having the tab open
close the Supermon tab and it stops

the other 2 systems on our network never exhibited this behavior again after the first episodes
I see lots of IRLP bits and pieces in the Supermon config files, and I tried excising some of it
but did not yet find the offender
to be honest I would like to be IRLP free

de k9wkj

I think you have but a few choices, but perhaps there are others that I have not given consideration.

You can start scanning files for the mentioned text

grep -Rnw '/path/to/search/' -e 'patterntext'

    -r or -R is recursive ; use -R to search entirely
    -n is line number, and
    -w stands for match the whole word.
    -l (lower-case L) can be added to just give the file name of matching files.
    -e is the pattern used during the search

Find where it is loading the call from.

Or you can backup your conf files, reload the software and try it again, perhaps on a new SD card so you can quickly revert if you need to…

You might first look in the system and user cron for it being called.

And if you still have issues, as I said before, I would satisfy it by making the file name it is looking for to stop the error message with a full permission set till you figure this out and that ‘might’ keep your log from exploding in the wait…

touch  /tmp/irlpdata.txt.gz