Seeing what's happening in library updates

[Tim Curtis]

That is very slow.

Slow throughput can be tricky to debug but the general approach is to first isolate the bottleneck by looking at performance statistics on the three main components: client, server, and network.

- What audio format(s) are the files?
- Look at the output of ifconfig and ifconfig commands to get a sense of whether there are network errors, retransmits, collisions etc.
- Run a ping test from the client to the server to see what the round trip times look like.

Virtually all .MP3

The one that works better:

eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.88.22  netmask 255.255.255.0  broadcast 192.168.88.255
        inet6 fe80::70eb:3ba0:88b2:7a08  prefixlen 64  scopeid 0x20<link>
        ether b8:27:eb:cb:9f:5b  txqueuelen 1000  (Ethernet)
        RX packets 44146047  bytes 41117865983 (38.2 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 14068122  bytes 1440127724 (1.3 GiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

Sample ping:

64 bytes from 192.168.88.59: icmp_seq=1 ttl=128 time=7.72 ms
64 bytes from 192.168.88.59: icmp_seq=2 ttl=128 time=4.81 ms
64 bytes from 192.168.88.59: icmp_seq=3 ttl=128 time=6.49 ms

The one that is slower:

eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.88.23  netmask 255.255.255.0  broadcast 192.168.88.255
        inet6 fe80::41bf:6861:fe07:f2a8  prefixlen 64  scopeid 0x20<link>
        ether b8:27:eb:9f:92:da  txqueuelen 1000  (Ethernet)
        RX packets 301189  bytes 316485271 (301.8 MiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 116028  bytes 14561001 (13.8 MiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

(not up for nearly as long...)

64 bytes from 192.168.88.59: icmp_seq=1 ttl=128 time=5.14 ms
64 bytes from 192.168.88.59: icmp_seq=2 ttl=128 time=5.11 ms
64 bytes from 192.168.88.59: icmp_seq=3 ttl=128 time=5.14 ms

I don't have the ifconfig for the time it was refreshing its library.

The long ping times suggest a possible network bottleneck. Ping times between two devices connected via Gbit Ethernet home runs to a Router are typically < 0.5ms avg and should have 0% packet loss.

Example:

Code:

pi@trx:~ $ ping moode-dev64
PING moode-dev64.home (192.168.1.134) 56(84) bytes of data.
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=1 ttl=64 time=0.820 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=2 ttl=64 time=0.327 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=3 ttl=64 time=0.330 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=4 ttl=64 time=0.382 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=5 ttl=64 time=0.424 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=6 ttl=64 time=0.325 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=7 ttl=64 time=0.322 ms
64 bytes from moode-dev64.home (192.168.1.134): icmp_seq=8 ttl=64 time=0.321 ms
^C
--- moode-dev64.home ping statistics ---
8 packets transmitted, 8 received, 0% packet loss, time 7011ms
rtt min/avg/max/mdev = 0.321/0.406/0.820/0.160 ms
pi@trx:~ $