Tablo Quad iperf3 performance not quite Gigabit?

Support & Troubleshooting
1

I recently purchased a Tablo Quad that I have placed on my hardwired Gigabit network. I recently discovered (through these forums) that it includes an iperf3 server. Running iperf3 tests from my Win10 machine and a Raspberry 4 show that it doesn’t quite run at Gigabit speeds. Additionally, I’m seeing non-zero values in iperf3’s ‘Retr’ field which seems suspect.

I’ve tried some minimal swapnostics by changing the Ethernet cable to a known, good CAT6 and have also tried a different port on my Ethernet switch. The performance characteristics have not changed.

I’m curious what others are seeing. It’s not exactly a deal breaker for me, but it is somewhat annoying that it’s not actually testing at the advertised speeds.

Here are my tests:

IP 192.168.0.53 = Tablo
IP 192.168.0.38 = Raspberry Pi 4
IP 192.168.0.102 = Win10 machine

RPi4 -> Tablo

pi@naspi:~ $ iperf3 -c 192.168.0.53
Connecting to host 192.168.0.53, port 5201
[  5] local 192.168.0.38 port 53686 connected to 192.168.0.53 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  41.0 MBytes   344 Mbits/sec   17    313 KBytes
[  5]   1.00-2.00   sec  40.3 MBytes   338 Mbits/sec    4    310 KBytes
[  5]   2.00-3.00   sec  40.4 MBytes   339 Mbits/sec   16    307 KBytes
[  5]   3.00-4.00   sec  40.1 MBytes   337 Mbits/sec   33    294 KBytes
[  5]   4.00-5.00   sec  40.1 MBytes   336 Mbits/sec   23    300 KBytes
[  5]   5.00-6.00   sec  40.3 MBytes   338 Mbits/sec   13    294 KBytes
[  5]   6.00-7.00   sec  40.3 MBytes   338 Mbits/sec   20    300 KBytes
[  5]   7.00-8.00   sec  39.6 MBytes   332 Mbits/sec   18    215 KBytes
[  5]   8.00-9.00   sec  40.4 MBytes   339 Mbits/sec    0    314 KBytes
[  5]   9.00-10.00  sec  40.3 MBytes   338 Mbits/sec   19    264 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec   403 MBytes   338 Mbits/sec  163             sender
[  5]   0.00-10.00  sec   402 MBytes   337 Mbits/sec                  receiver

iperf Done.

RPi4 -> Win10

pi@naspi:~ $ iperf3 -c 192.168.0.102
Connecting to host 192.168.0.102, port 5201
[  5] local 192.168.0.38 port 58302 connected to 192.168.0.102 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec   111 MBytes   929 Mbits/sec    0    365 KBytes
[  5]   1.00-2.00   sec   113 MBytes   945 Mbits/sec    0    365 KBytes
[  5]   2.00-3.00   sec   113 MBytes   948 Mbits/sec    0    365 KBytes
[  5]   3.00-4.00   sec   113 MBytes   948 Mbits/sec    0    365 KBytes
[  5]   4.00-5.00   sec   113 MBytes   948 Mbits/sec    0    365 KBytes
[  5]   5.00-6.00   sec   113 MBytes   949 Mbits/sec    0    365 KBytes
[  5]   6.00-7.00   sec   113 MBytes   947 Mbits/sec    0    365 KBytes
[  5]   7.00-8.00   sec   113 MBytes   950 Mbits/sec    0    365 KBytes
[  5]   8.00-9.00   sec   112 MBytes   944 Mbits/sec    0    365 KBytes
[  5]   9.00-10.00  sec   113 MBytes   951 Mbits/sec    0    365 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  1.10 GBytes   946 Mbits/sec    0             sender
[  5]   0.00-10.04  sec  1.10 GBytes   941 Mbits/sec                  receiver

Win10 -> Tablo

C:\apps\iperf3.6_64bit>iperf3 -c 192.168.0.53
warning: Ignoring nonsense TCP MSS 0
Connecting to host 192.168.0.53, port 5201
[  5] local 192.168.0.102 port 53342 connected to 192.168.0.53 port 5201
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-1.00   sec  40.6 MBytes   341 Mbits/sec
[  5]   1.00-2.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   2.00-3.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   3.00-4.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   4.00-5.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   5.00-6.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   6.00-7.00   sec  43.1 MBytes   362 Mbits/sec
[  5]   7.00-8.00   sec  43.1 MBytes   361 Mbits/sec
[  5]   8.00-9.00   sec  43.2 MBytes   363 Mbits/sec
[  5]   9.00-10.00  sec  43.1 MBytes   362 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-10.00  sec   429 MBytes   360 Mbits/sec                  sender
[  5]   0.00-10.00  sec   428 MBytes   359 Mbits/sec                  receiver

iperf Done.
2

If you’re playing a single stream… that’s as fast as it “plays”. The tablo uses/is an HTTP server just as any on the internet, it streams the recordings via HLS or HTTP Live Streaming, so it doesn’t send an entire recording to your app/device.

If you have several, you might see more traffic. But if it can perform over WiFi just the same, you’ll unlikely see it using the entire 1000/Mbs used.

3

Appreciate the reply. I understand and realize all of that. The point of the post was not so much that Tablo Quad is not “fast enough” (I have zero troubles streaming) but rather, “not fast as spec’d” – at least according to the performance tool that it has built-in.

I’m curious if any others in the community are seeing similar performance numbers.

4

I don’t have a Gigabit port on my Tablo DVR but I’m guessing that the results you’re seeing are most likely due to a limitation with the processing unit rather than with the Gigabit port.

1 Like
5

Yea, that’s kind of what I was trying to convey. It just doesn’t have the capacity to transmit that much data.

What are you “making” it do to max out it’s bandwidth? Just because it has gigabit throughput doesn’t mean it can use it all.

If it can perform at full capacity over WiFi, again, what are you making it do to make it fill 1000Mbps? It’s only going to stream as fast as X number of devices can process their stream… isn’t the same as

6

And remember that the Tablo is designed to work with a 10/100 network connection. I am sure the Tablo folks added a gigabit card to the newest Tablos because the price difference was so minimal (if they even make 10/100 port cards anymore).

1 Like
7

Maybe not gigabit speed but the Quad is faster then the original 4-tuner. However I don’t know what else besides the nic changed between the quad and 4-tuner.

4-tuner 4 streams.
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 29.3 MBytes 24.6 Mbits/sec 0 sender
[ 5] 0.00-10.00 sec 29.1 MBytes 24.5 Mbits/sec receiver
[ 7] 0.00-10.00 sec 27.7 MBytes 23.2 Mbits/sec 0 sender
[ 7] 0.00-10.00 sec 27.5 MBytes 23.0 Mbits/sec receiver
[ 9] 0.00-10.00 sec 28.7 MBytes 24.1 Mbits/sec 0 sender
[ 9] 0.00-10.00 sec 28.6 MBytes 24.0 Mbits/sec receiver
[ 11] 0.00-10.00 sec 27.2 MBytes 22.8 Mbits/sec 0 sender
[ 11] 0.00-10.00 sec 27.1 MBytes 22.7 Mbits/sec receiver
[SUM] 0.00-10.00 sec 113 MBytes 94.7 Mbits/sec 0 sender
[SUM] 0.00-10.00 sec 112 MBytes 94.2 Mbits/sec receiver

Quad 4 streams
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 107 MBytes 89.4 Mbits/sec 0 sender
[ 5] 0.00-10.00 sec 105 MBytes 88.1 Mbits/sec receiver
[ 7] 0.00-10.00 sec 106 MBytes 89.1 Mbits/sec 0 sender
[ 7] 0.00-10.00 sec 105 MBytes 88.1 Mbits/sec receiver
[ 9] 0.00-10.00 sec 106 MBytes 88.8 Mbits/sec 0 sender
[ 9] 0.00-10.00 sec 105 MBytes 87.8 Mbits/sec receiver
[ 11] 0.00-10.00 sec 102 MBytes 85.8 Mbits/sec 0 sender
[ 11] 0.00-10.00 sec 101 MBytes 85.1 Mbits/sec receiver
[SUM] 0.00-10.00 sec 421 MBytes 353 Mbits/sec 0 sender
[SUM] 0.00-10.00 sec 416 MBytes 349 Mbits/sec receiver

1 Like
8

@andersonas25 Good to know! Thanks for providing your test results.

9

I know they have different tuners and processors. Just asking, are you streaming the same 4 shows to the same for devices? And the quality of the shows are the same?

If so, bandwidth isn’t necessarily a bottle neck, if less than “as spec’d”

10

By four streams I meant iperf3 parallel client streams, not video.

I guess it is also possible the old 4-tuner was bottlenecked by the 10/100 nic.

11

and we can back up a bit and note, iperf3 measures throughput across an entire network - not specifically a device’s bandwidth.

I guess I didn’t completely understand, you’re not truly measuring a device with that tool - it’s an entire network tool.

12

Wikipedia has a good overview of iperf operation: https://en.wikipedia.org/wiki/Iperf

13

That’s a nice easy to read article. Nowhere in iperf3’s docs man/help pages use the phrase -

to measure the throughput between the two ends in one or both directions.

https://iperf.fr/

iPerf3 is a tool for active measurements of the maximum achievable bandwidth on IP networks.

14