Strange squid process behavior

no nothing in common except that everyone has installed:
guardian
clamav
squidclamav

But before January I didn’t notice these anomalies, for example my graph of processes in memory regarding squid has always stood at 200 MB on average in the last year, only since January has it gone crazy.

I noticed, because on all four systems, navigation was slow and sometimes the fact was reported that some sites that maybe worked a few minutes before, the browser reported that they weren’t reachable.

I logged into ipfire, cleared the cache and everything worked again.

Hi,
I didn’t mean the IPFire systems, but the attached networks.
Squid handles the web requests of the LAN. Without these there will no problem. :wink:
Have you tried to switch off squidclamav? I don’t remember an issue with this at the moment, but you never know.

The networks are different, I have about 30 home automation devices, but they were there even in December, one is an office with 2 people and the other is my parents’ firewall.

I have squidclamav deactivated from 10 days just to understand if it was him

Maybe this and this helps.

There are other topics about clamav found by searching.

Thanks, but does clamav have any logs?
I’ve looked in /var/log but can’t find anything

In the wiki page link I gave earlier in this thread related to the cache it mentions that the cache memory value is not a fixed limit but under high load squid can override the value.

There is also a note that the cache memory value is not the total memory footprint of the cache.

1 Like

I believe that you have to grep in /var/log/messages or you can look in the wui menu Logs - System Logs and then select ClamAV in the Section: drop down box and then press the Update button.

1 Like

Could it be the content filter causing this problem?

I’m telling you this because I tried disabling it on one of the firewalls and it’s been 4 days since the request storm last appeared.

Good morning,
we are on the third day with the url filter disabled and the connection storm has not yet reoccurred, all four firewalls involved are behaving the same way, I keep my fingers crossed.

1 Like

Update on the situation,
it is now 10 days that on all 4 firewalls, with the content filter deactivated, the strange storm of requests does not appear.

Now I’d like to know why this happens.
I’ll try to reactivate the filter to see if everything comes back.

3 Likes

Good morning,
it’s been 20 days now that with the content filter deactivated, the problem of the storm of requests no longer occurs, on mine the only peak is when I watch the amazon prime and disney plus services.
Also for the other firewalls everything is ok.
All are updated to CU 173, I’ll try to re-enable content filtering on mine to see what happens.

Hi @sky7176 ,

I seem to have the same problem as you :frowning:

https://community.ipfire.org/t/lots-of-none-none-000-0-error-transaction-end-before-headers-hier-none-error-messages-in-the-squid-log/9548

Did the problem reappear at your place ?
Thanks

HI,
sorry for the delay in the response.
No, it hasn’t happened to me in over a month.
Other than what I indicated in the post, I did nothing, as it appeared it disappeared.
I’ve never been able to give myself an explanation.
DOS to my four firewalls?