my download speeds as per ookla speed tests. I test 4 different servers with r50176 then reflash the router and retest the same servers with the r50146 within 10 minutes. I have seen this same situation sporadically over the past year. I try the new flashes and test the Wireguard throuput and have seen discrepancies several times I just go by the numbers and if they are correct 7-8 points on my download speeds is quite a bit considering my basic plan at 50 down and 10 up.thank you
Posted: Wed Sep 21, 2022 5:13 Post subject: Max DHCP users defaulted to 190, causing DHCP to fail
Router/Version: R7000
File/Kernel: 50176
Previous/Reset: Reset to defaults
Mode/Status: No changes from defaults
Issues/Errors: DHCP fails. Noticed Max DHCP users defaulted to 190 and end ip address was next subnet (odd choice of end value, but not necessarily incorrect). After lowering max dhcp users to 10 and restarting, DHCP is no longer failing.
Posted: Wed Sep 21, 2022 14:53 Post subject: Re: Max DHCP users defaulted to 190, causing DHCP to fail
gguruusa wrote:
Issues/Errors: DHCP fails. Noticed Max DHCP users defaulted to 190 and end ip address was next subnet (odd choice of end value, but not necessarily incorrect). After lowering max dhcp users to 10 and restarting, DHCP is no longer failing.
If by chance your start IP was .100, then that makes sense. The default start IP was changed to start at .64 a while back, and the default number of leases 190. Some exceptions apply:
The bufferbloat tests are all unusable.
Not that the tests are bad in themselves, but they use some far away server and measure some bufferbloat on the internet and not the bufferbloat of the router.
tcpdump reveals that waveform uses cloudflare infrastructure (and measures absolute nonsense from time to time)
The best way to measure bufferbloat is to start a download and ping a nearby high availability server in very short intervals.
This gives much more useful results than any server in timbuktu.
So don't rely on any bufferbloat test sites they all measure garbage
This is already covered, but any speed test site's reliability is highly dependent on server locations relative to yours.
Speedtest.net is good for me, so many local servers, new feature idle, download and upload latency measurements.
Waveform works well for me, but again only because I am lucky to be close enough to their servers. nPerf is alright.
dslreports RIP. This is one is pretty neat, https://testmy.net/latency and of course set up your own as shown above.
Joined: 16 Jun 2022 Posts: 16 Location: Dallas, TX
Posted: Mon Sep 26, 2022 2:35 Post subject:
Firmware: DD-WRT v3.0-r50146 std (09/10/22)
Router/Version: ASUS RT-AC5300
Kernel: Linux 4.4.302-st25 #7386 SMP Sat Sep 10 15:26:03 +07 2022 armv7l
Previous: DD-WRT v3.0-r49626 std (08/03/22)
Mode/Status: Wired and Wireless (1x 2.4 GHz, 2x 5 GHz) as Router (with DHCP), Switch and AP. DNS is offloaded to PiHole.
Uptime: 19 min, Prior to Upgrade: 44 days
Reset: No
Temperatures: CPU 55.2 °C / WL0 43.5 °C / WL1 44.6 °C / WL2 48.8 °C
Issues/Errors: All good, snug as a bug in a rug.
Aiming for monthly update cadence for DD-WRT now that hardware temperatures are stable.