[Groma]

DIR-825 revB

04-14-2019-r39508 - 08-25-2019-r40784


Mem: 45204K used, 16544K free, 0K shrd, 11732K buff, 11652K cached
CPU: 0.1% usr 0.1% sys 0.0% nic 99.4% idle 0.0% io 0.0% irq 0.1% sirq
Load average: 4.03 4.02 4.05 6/46 8081
3281 1 root S 5108 8.2 0 0.0 httpd -p 80

GUI not work.


kill 3281

in 20 minutes

Mem: 42820K used, 18928K free, 0K shrd, 11748K buff, 11684K cached
CPU: 0.0% usr 0.1% sys 0.0% nic 99.8% idle 0.0% io 0.0% irq 0.0% sirq
Load average: 0.04 0.12 1.18 2/41 8902
8099 1 root S 4408 7.1 0 0.0 httpd -p 80

[kernel-panic69]

I'm curious what you have configured. My antique Broadcom isn't nearly as loaded most of the time: Was this just after flashing an upgrade?

[Groma]

on DD-WRT v3.0-r38240 std (01/08/19), everything is fine. then went these problems throughout the branch. the settings were not changed. all standard.

the same average load as you I have happens after reboot.

[kernel-panic69]

How long was it after you flashed the upgrade that the load numbers went through the roof? I don't think it is uncommon, some builds / devices have more issue than others, I suspect. I had a weird issue many builds back with load going screwy until I rebooted again, so hmm.

[mrjcd]

I install lots of DD-WRT builds on the EA8500 and hardly ever do a reset and ever so often the webif goes to heck soon after an install.
Just happened a couple hours ago after installing a BS test build.
A simple power cycle (switch off 2 sec then back on) fixed it and no httpd trobles at all now.

[Groma]

here since r38240 is a constant problem.
this is not only me, it is observed on other devices.

[kernel-panic69]

I generally don't do a reset either, since I started using an addition to my startup script that seems to alleviate the usual issues. The variables don't change all that much over time, but. Someone mentioned to me that it's not a bad idea to reboot again after the post-flash reboot, just in case, after checking for issues.

One thing that makes me scratch my head and boggles my mind is some things are bound by defines, and not cases in the code. I think you could alleviate a lot of per-device issues if it were case $(devicename): in some sections of the code, but I don't know what the hell I'm talking about

[Groma]

included httpd -S, see.

[kernel-panic69]

httpd -S ... doesn't that have a known issue to increase load and lock up still? I think I recall reading about that somewhere.

[tatsuya46]

[Wildlion]

Groma,

I hate to say it but mine (Dir-825 rev B) is working fine:


Mem: 24748K used, 37000K free, 0K shrd, 2604K buff, 7756K cached
CPU: 0.7% usr 0.3% sys 0.0% nic 98.8% idle 0.0% io 0.0% irq 0.0% sirq
Load average: 0.01 0.02 0.05 2/34 10531
15419 1 root S 4344 7.0 0 1.0 httpd -p 80

this is on 07-04-2019-r40189 and has an uptime of over 54 days.



So it makes me think that it is a customization or hardware on your side. How is yours deployed? What settings do you have?

[Groma]

the inclusion of httpd-S showed that at the same settings the router lasted one day without reboots.

Firmware: DD-WRT v3.0-r40784 std (08/25/19)
Time: 07:49:36 up 1 day, 4:49, load average: 0.02, 0.04, 0.04

so the problem is present in httpd-p 80 and absent in httpd-S.