00:23
<
slh >
\x: Xiaomi be6500 is probably quite similar (~160 EUR inkcl 19% VAT delivered from aliexpress), if the seller specs are to be believed
00:50
<
\x >
slh: acwifi has a teardown of be6500, sadly no expandability
00:51
<
\x >
but ofcourse the question here is, "does usb matter?" for me it does, im one of those "no usb no buy"
00:52
<
\x >
castiel652: thats good, ehemm any /proc/cpuinfo dump? they likely are using recycled kryo mobile cores again
01:01
thejoker8814 is now known as Guest1681
01:01
thejoker8814 has joined #openwrt-devel
01:07
<
mrkiko >
however, ipq4xx drivers in FB7530 do exhibit broken behaviours sometimes, atleast that is on the unit I am using with my setup
01:07
<
mrkiko >
*ethernet-related drivers*
01:07
<
mrkiko >
interposing a switch between the device and the rest of the network fixed issue for me
01:09
Guest1681 has quit [Ping timeout: 480 seconds]
01:09
goliath has quit [Quit: SIGSEGV]
01:20
Emantor has joined #openwrt-devel
01:30
<
slh >
\x: thanks, USB is something I'd really like to have - but sadly not a hard requirement (as that would disqualify half of the devices on the market)
01:33
<
slh >
SOC, RAM size, number of ethernet ports and their specs, as well as 6 GHz capability (yeah, that's not to be found in device for the Chinese market) take precedence over USB. it's great to have it though - and I do miss it, rarely, but when I do, I really miss it
01:58
minimal has quit [Quit: Leaving]
02:00
tSYS has quit [Quit: *squeak*]
02:00
tSYS has joined #openwrt-devel
02:02
<
mrkiko >
slh: out of curiosity - what was your use-case for usb on the router?
02:08
<
slh >
mrkiko: two things, printer(s) and some small storage space, be it for logs (rrd/ luci statistics) and a bit more custom, a small tftpd-hpa instance to serve the firmware- and configuration to my SIP phone(s) (Cisco cp-7962g, cp7960g)
02:10
<
slh >
small stuff: 15.8M /srv/tftp/, 1.7M /srv/rrd/, 756.0K /srv/nlbwmon/
02:40
<
mrkiko >
slh: got it
03:34
<
\x >
damn mrkiko did not ask me
03:38
<
\x >
would be nice though if nlbwmon supported remote logging amirite
03:56
gch981213 has joined #openwrt-devel
04:00
SirLouen has quit [Quit: Hasta Luego, Lucas]
04:00
SirLouen has joined #openwrt-devel
04:06
<
slh >
I've never seen the point of mounting remote filesystems on routers, the router should always work - without depending on the rest of the network
04:25
<
mrkiko >
slh: well, it enables fun stuff I think
04:25
<
mrkiko >
slh: there shouldn't always be a point in any case
04:25
<
mrkiko >
\x: I can't see images easily but nice to know
04:36
rua has quit [Quit: Leaving.]
05:11
vincejv has quit [Quit: Bye bye! Leaving for now...]
05:11
vincejv has joined #openwrt-devel
05:45
sestowner has joined #openwrt-devel
06:01
johnf has joined #openwrt-devel
06:50
rua has joined #openwrt-devel
06:51
antonkh has joined #openwrt-devel
06:54
<
svanheule >
mirko: I've looked at the fan monitor line in the past, but never managed to figure out how it worked
06:55
<
svanheule >
mirko: it could also be that they're passing it through a low-pass filter as a "stall alarm" more than a direct RPM measurement
08:35
Borromini has joined #openwrt-devel
08:53
Borromini has quit [Ping timeout: 480 seconds]
08:59
Borromini has joined #openwrt-devel
09:13
<
aparcar >
svanheule aren’t you the switch guru?
10:00
Borromini has quit [Ping timeout: 480 seconds]
10:15
Borromini has joined #openwrt-devel
10:42
castiel652 has joined #openwrt-devel
11:25
<
mrnuke >
svanheule: I have an init dump from a Realtek PSE (the actual RTL8..., not broadcom)
11:29
<
mrnuke >
Specifically, it's the SoC to MCU lines UART dump (TX/RX)
11:32
<
castiel652 >
\x: sure I can check that on Monday. You can look up NWA130BE if you are interested in the actual product. ;)
11:40
Borromini has quit [Quit: Lost terminal]
11:44
rua has quit [Quit: Leaving.]
12:12
<
\x >
castiel652: my suspicion is that qualcomm is puting again kryo 200 silver cores on them, which is closer to a73 than a53
12:12
<
\x >
they did it on ipq60xx and 50xx
12:12
<
\x >
but they marketed it as "cortex-a53"
12:23
sestowner has quit [Ping timeout: 480 seconds]
12:33
goliath has joined #openwrt-devel
13:33
castiel652 has quit [Quit: Leaving]
13:58
dangole has joined #openwrt-devel
14:50
rua has joined #openwrt-devel
15:00
mentalow has quit [Quit: :]]
15:01
mentalow has joined #openwrt-devel
15:23
rua has quit [Quit: Leaving.]
15:32
rua has joined #openwrt-devel
15:52
<
mirko >
svanheule: interesting, internet says switch just continues working when fan is disconnected
15:53
<
mirko >
figured maybe it's not even used, but there's electronic components on the lane - however couldn't trace it to the SoC
17:38
mrkiko has quit [Quit: leaving]
18:15
<
antonkh >
@mirko how many wires does the fan have?
19:01
mark22k has joined #openwrt-devel
19:07
<
mrnuke >
the frogging global context in realtek-poe is killing me!
19:11
<
mirko >
antonkh: 3, VCC, GND, RPM
19:20
<
mirko >
mrnuke: is there actually a way of reading out how much power a port draws?
19:21
<
mrnuke >
mirko: ubus call poe info
19:23
<
mirko >
mrnuke: uh, nice, thanks
19:24
<
mrnuke >
mirko: you gotta read the documentation that I never wrote! :p
19:25
<
mirko >
mrnuke: classic openwrt style :)
19:27
<
mirko >
mrnuke: can i also enable/disable PoE without editing config files and restarting network?
19:28
<
mirko >
next question: can i classify/prioritise ports re power consumption?
19:29
<
mrnuke >
mirko: see "ubus -v list poe" . If I tried to remember the syntax I'd be lying
19:29
<
mrnuke >
you can try the "priority"field in the /etc/config/poe, but we never tested if that actually works
20:07
<
svanheule >
aparcar: I suppose some might be inclined to call me that
20:08
<
svanheule >
mirko: when debugging I often leave the fans disconnected. You're right that the devices don't really seem to care
20:08
<
aparcar >
svanheule: did you ever run openwrt on this ubiquiti aggregation switch?
20:09
<
svanheule >
aparcar: no, I didn't. That was Birger's device
20:10
<
aparcar >
svanheule: all right. I ordered one if you want remote access to play with it let me know
20:10
<
aparcar >
looks like quite a cool device except for the silly screen
20:11
<
svanheule >
aparcar: ack. Although if I find time I would probably prioritize fixing the 1st gen (GbE) switches
20:12
<
svanheule >
I have the Unifi Switch Lite 16 PoE. It's pretty well built IMHO, but they're really useless without their controller
20:15
<
aparcar >
I think david aka blocktrron looked into some unifi controlling stuff, not sure if that device specifically
20:16
<
svanheule >
mrnuke: let me have a look. does that match the control sequences of the BCM ones well?
20:16
<
schmars[m] >
blocktrron did usw-flex which works nicely
20:16
<
aparcar >
schmars: also with PoE?
20:16
<
schmars[m] >
yeeah, really nice device
20:16
<
schmars[m] >
different poe chip than the realteks though
20:16
<
svanheule >
schmars[m]: is that the mt7621 switch?
20:17
<
aparcar >
blocktrron: why did yo never tell me 😄 did we struggle with that very device the last time we met?
20:19
<
aparcar >
btw what about the flex mini?
20:24
<
blocktrron >
aparcar: flex mini uses nuvoton MCU for MGMT
20:25
<
blocktrron >
Newer unifi ultra use esp32 for management. Cutting costs where possible
20:26
<
aparcar >
so the flex mini is hard to support?
20:27
<
aparcar >
is poe++ just passive 48v?
20:28
<
schmars[m] >
nah it's active - 802.3bt
20:28
<
schmars[m] >
poe = 802.3af, poe+ = 802.3at, poe++ = 802.3bt
20:30
<
schmars[m] >
ubiquiti has some weird segmentation going on, where some of their stuff is active, and some is passive
20:40
<
mrnuke >
svanheule: no way. It's completely different
20:40
<
mrnuke >
svanheule: currently hacking the golab contexts out of realtek-poe so I can add a new dialect
20:40
<
mrnuke >
s/golab/global/
20:51
<
mrnuke >
well, obviously, they reused some ideas, but the command IDs are all different. I am also only seeing 1-port commands being used. None of the fancy poet/4_port stuff
21:24
<
russell-- >
mrnuke++
22:44
goliath has quit [Quit: SIGSEGV]
22:46
<
Mangix >
hrm marko's not here
22:47
<
russell-- >
mrnuke: i have a zyxel gs1900-10hp v2, which doesn't work with realtek-poe, but the PSE chips have heatsinks, so i haven't ID'd them
23:02
fakuivan has quit [Quit: No Ping reply in 180 seconds.]
23:02
fakuivan has joined #openwrt-devel
23:05
<
mrnuke >
do you get a bunch of "Invalid port status packet (port=97)" from realtek-poe?
23:26
<
russell-- >
mrnuke: i think so, something like that was filling up my syslogs so i disabled the daemon
23:27
kwz has joined #openwrt-devel
23:27
<
russell-- >
Mar 24 00:01:39 10.11.20.5 realtek-poe: Invalid port status packet (port=97)
23:39
<
mrnuke >
russell--: Welcome to the "unrealtek could have just kept the same damn protocol, but they didn't" clob
23:40
<
mrnuke >
s/clob/club/
23:52
<
russell-- >
lol, yeah.