B3, Logitech SqueezeServer

[mmmark]

Am having a very annoying problem with my Squeezeboxes occasionally dropping their connection to the B3. Happens to all of them simultaneously (I have three dotted about the house) so the server is definitely the item at fault. In the daemon log I see the following:

Code: Select all

Jun 24 13:47:02 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: deauthenticated due to local deauth request
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: authenticated
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: associated (aid 1)
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd RADIUS: starting accounting session 53A7BCFB-0000007F
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd WPA: pairwise key handshake completed (RSN)
Jun 24 13:47:10 b3 bubba-aptd[1543]: Resetting daemon
Jun 24 13:47:54 b3 bubba-networkmanager: Server timed out, terminating
Jun 24 13:47:54 b3 bubba-networkmanager: Daemon terminating
Jun 24 13:47:54 b3 bubba-networkmanager: Shutting down

It looks as though bubba-aptd resets itself for some reason, and this causes the bubba-network manager to shut down. Is this expected?

Syslog shows about the same:

Code: Select all

Jun 24 13:47:02 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: deauthenticated due to local deauth request
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: authenticated
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd IEEE 802.11: associated (aid 1)
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd RADIUS: starting accounting session 53A7BCFB-0000007F
Jun 24 13:47:09 b3 hostapd: wlan0: STA 00:04:20:12:5b:fd WPA: pairwise key handshake completed (RSN)
Jun 24 13:47:10 b3 bubba-aptd[1543]: Resetting daemon
Jun 24 13:47:54 b3 bubba-networkmanager: Server timed out, terminating
Jun 24 13:47:54 b3 bubba-networkmanager: Daemon terminating
Jun 24 13:47:54 b3 bubba-networkmanager: Shutting down
Jun 24 13:50:01 b3 /USR/SBIN/CRON[8132]: (root) CMD (/etc/init.d/dovecot status >/dev/null 2>&1 || /etc/init.d/dovecot restart)

Sometimes the squeezeboxes pick up the connection to the server again after a few minutes, other times only restarting the b3 brings everything back to life. There's nothing in the SqueezeServer logs that corresponds to the time of the cutouts.

Any assistance would be much appreciated.

[Gordon]

I don't think bubba-aptd is related to the issue. And bubba-networkmanager isn't actually a daemon but some kind of helper application to identify networking parameters. Obviously something is wrong there in your B3, causing it to time out. Are you running the B3 in server only mode?

I actually have three Squeezebox players myself, and Squeezebox Controller as a fourth device. I did use to have similar issues as you describe, but in my case it was never Squeezebox server (LMS) that caused the troubles. If a player lost connection, I could never authenticate against the wifi network anymore and I would need to restart the player in question. I can't remember if the issues would occur with all players every time, but it definitely did some times.

I'm unsure why it hasn't occurred for a long time now. What I did notice is that the Squeezebox devices spit out a lot of broadcast messages and that also caused me issues with the tiny CPU in the tablet PC I use in the living room. I have totally reconfigured networking on the B3, running three different SSID's on wifi and two VLAN's wired. On one VLAN I have my Squeezebox receiver which is out of wifi range and this one is bridged with an unencrypted wifi network to which the other three Squeezebox devices connect. I like to think that dropping WPA encryption did the trick here.

[johannes]

I have similar issues sometimes, I can't be sure but I think it's just wifi wipe-outs that temporarily cause the connection to fail, and it seems like the squeezebox clients don't retry connecting. After squeezebox restarts it works fine for me.

Wifi wipe-outs in my house are caused by turning my microwave on, so it could have lots of reasons, not neccesarily related to these products.

And of course it could also be the B3 failing, but the log is normal and I don't think that's related.

[RandomUsername]

I used to get a similar problem with my Sky (satellite TV) box. I ended up creating a cron job that restarts hostapd (the wifi daemon) every morning and so far it's all been OK.

[mmmark]

Thanks all for your replies. I'll try the cron job route first and see how I go.

It does seem strange to me that the connections are so fragile, but it generally doesn't hurt to restart every now and again.

[Gordon]

Me and my big mouth

As it happened, last night SB radio could not find the library any more. Although wifi connection seemed okay, I did a full reset of the device to no result. Checking this morning I could also not get into the web console for LMS and so I had to restart the service. Apparently it had gone stuck during a media scan a while back and seized operating just now.

I also had a bit of an epiphany, remembering that I changed some other stuff as well. I wrote it down in this thread a few months ago.