add total latency to perf charts

[totaam]

Issue migrated from trac ticket # 2381

component: tests | priority: major | resolution: needinfo

2019-08-09 13:24:42: antoine created the issue

---

As part of #619 and #2130, it would help to be able to see how long screen updates take to land on the client's screen.
In the same way that we have the "overall latency" on the session info graphs.

[totaam]

2019-08-19 17:38:36: antoine changed status from new to assigned

[totaam]

2019-08-19 17:38:36: antoine edited the issue description

[totaam]

2019-08-20 11:25:36: antoine changed status from assigned to new

[totaam]

2019-08-20 11:25:36: antoine changed owner from antoine to smo

[totaam]

2019-08-20 11:25:36: antoine commented

---

r23539 records two new attributes in xpra info:

• client.damage.frame-total-latency: how long it takes from the moment we get a damage event until we get the draw acknowledgement packet from the client, that's everything we do - which can be summarized as: batch delay + compression + sending + network latency + receiving + updating the window buffer + sending ack + network latency (again) + receiving ack
• client.damage.client-latency: same, minus the connection latency (note: this latency includes the cost of the protocol layer - which is small in most but not all cases..)

As of r23541, these new attributes are now exposed in the perf test data as:

• "Frame Total Latency"
• "Client Frame Latency"

@smo: does that work?
Maybe re-run just NODELAY=0 + CORK=0 vs default settings.

[totaam]

2019-08-20 15:48:23: smo commented

---

Added these to charts in r23546

[totaam]

2019-08-21 01:41:20: smo uploaded file test_nodelay0cork0.tar.gz (238.0 KiB)

nodelay0 cork0 vs defaults

[totaam]

2019-08-21 01:42:13: smo commented

---

Here are the results attached to the ticket.

Something looks off to me what do you think?

[totaam]

2019-08-21 06:16:20: antoine uploaded file test-client-frame-latency-dump.patch (0.8 KiB)

debug client frame latency

[totaam]

2019-08-21 06:20:48: antoine commented

---

Something looks off to me what do you think?
Yup. The values are completely wrong: they should usually be lower than 100ms (+network latency).
No idea why. The values are correct when I run xpra info | egrep "damage.frame-total|damage.client-latency" and the perf test code uses the same function as everywhere else for capturing the data:

add("", avg, "Frame Total Latency",             ["client.damage.frame-total-latency"])
add("", avg, "Client Frame Latency",            ["client.damage.client-latency"])

I've also run the tests and got sensible values out:

Frame Total Latency	 Client Frame Latency
34	34
43	43
45	45
32	32
77	77
63	63

Can you run again with the patch attached and see what you get? It will dump the values it finds for "Client Frame Latency" as it goes along.

[totaam]

2020-02-12 05:21:10: antoine changed status from new to closed

[totaam]

2020-02-12 05:21:10: antoine set resolution to needinfo