OLA Thread Scheduling - Revision history

Nomis52: /* Results */

2014-12-29T04:43:06Z

‎Results

Nomis52 at 03:38, 29 December 2014

2014-12-29T03:38:48Z

Nomis52 at 03:24, 29 December 2014

2014-12-29T03:24:35Z

Nomis52: /* Results */

2014-12-01T05:08:11Z

‎Results

Nomis52: /* Results */

2014-12-01T04:40:32Z

‎Results

Nomis52 at 20:51, 30 November 2014

2014-11-30T20:51:32Z

Nomis52 at 20:50, 30 November 2014

2014-11-30T20:50:12Z

Nomis52: Created page with "Since 0.9.4, OLA provides options to control the thread scheduling algorithms, when supported by the host kernel. Besides the default scheduling policy (SCHED_OTHER), OLA sup..."

2014-11-30T20:47:41Z

Created page with "Since 0.9.4, OLA provides options to control the thread scheduling algorithms, when supported by the host kernel. Besides the default scheduling policy (SCHED_OTHER), OLA sup..."

New page

Since 0.9.4, OLA provides options to control the thread scheduling algorithms, when supported by the host kernel.

Besides the default scheduling policy (SCHED_OTHER), OLA supports SCHED_RR and SCHED_FIFO. You can read about the differences on [http://www.linuxjournal.com/article/3910| The Linux Journal].

== Configuration ==

If you're on Linux you need to edit <tt>/etc/security/limits.conf</tt> and add:

<pre>
<username> hard rtprio 99
<username> soft rtprio 50
</pre>

Replace <username> with the user you run olad as.

== Running OLA ==

Start olad with <tt>--scheduler-policy</tt> and <tt>--scheduler-priority</tt>. You must provide both flags.

<pre>
$ olad --scheduler-policy fifo --scheduler-priority 10
</pre>

The --scheduler-priority flag must be a value less than or equal to the soft rtprio limit set in <tt>/etc/security/limits.conf</tt>.

== Results ==

@@ Line 51: / Line 51: @@
 From the tcpdump, we can find the time between sending the ArtNet packet and receiving the E1.31 packet, and then plot these on a histogram:
-{|
+{| class="wikitable"
-! Scenario
+! Scheduler
-! Mean
+! Mean (ms)
 ! Standard Deviation
-! 90%
+! 90 percentile (ms)
-! 99%
+! 99 percentile (ms)
-! 99.9 %
+! 99.9 percentile (ms)
-! Max
+! Max (ms)
 |-
 | Control
-| 0.0008808826
+| 0.880 ms
-| 5.64791e-05
+| 5.64791e-02
-| 0.000920000
+| 0.920 ms
-| 0.001018000
+| 1.018 ms
-| 0.001842056
+| 1.842 ms
-| 0.002252
+| 2.252 ms
 |-
 | Fifo
-| 0.0008681658
+| 0.868 ms
-| 2.819106e-05
+| 2.819106e-02
-| 0.000909000
+| 0.909 ms
-| 0.000958020
+| 0.958 ms
-| 0.001010026
+| 1.010 ms
-| 0.001132
+| 1.132 ms
 |}

← Older revision		Revision as of 03:38, 29 December 2014
Line 50:		Line 50:

	From the tcpdump, we can find the time between sending the ArtNet packet and receiving the E1.31 packet, and then plot these on a histogram:		From the tcpdump, we can find the time between sending the ArtNet packet and receiving the E1.31 packet, and then plot these on a histogram:
		+
		+	{\|
		+	! Scenario
		+	! Mean
		+	! Standard Deviation
		+	! 90%
		+	! 99%
		+	! 99.9 %
		+	! Max
		+	\|-
		+	\| Control
		+	\| 0.0008808826
		+	\| 5.64791e-05
		+	\| 0.000920000
		+	\| 0.001018000
		+	\| 0.001842056
		+	\| 0.002252
		+	\|-
		+	\| Fifo
		+	\| 0.0008681658
		+	\| 2.819106e-05
		+	\| 0.000909000
		+	\| 0.000958020
		+	\| 0.001010026
		+	\| 0.001132
		+	\|}
		+
		+	[[File:Thread-scheduling-control.png]]
		+
		+	[[File:Thread-scheduling-fifo.png]]

@@ Line 34: / Line 34: @@
 <pre>
-$  i=0; while :; do echo "$i"; i=$((i+1)); if [ $i == 256 ] ; then i=0; fi ; sleep 0.1; done | ola_streaming_client  -u 1
+$  i=0; while :; do echo "$i"; i=$((i+1)); if [ $i == 256 ] ; then i=0; fi ; sleep 0.025; done | ola_streaming_client  -u 1
 </pre>
@@ Line 40: / Line 40: @@
 <pre>
-tcpdump  -n '(port 5568 or port 6454)  and (dst host 239.255.0.1 or dst host 192.168.3.115)' -c 2000 -s0 -w /tmp/savefile-fifo
+tcpdump  -n '(port 5568 or port 6454)  and (dst host 239.255.0.1 or dst host 192.168.3.115)' -c 10000 -s0 -w /tmp/savefile-fifo
 </pre>
@@ Line 48: / Line 48: @@
 $ olad  --scheduler-policy fifo --scheduler-priority 10
 </pre>

@@ Line 40: / Line 40: @@
 <pre>
-tcpdump port 5568  or port 6454   -n  -s0 -w /tmp/savefile
+tcpdump  -n '(port 5568 or port 6454)  and (dst host 239.255.0.1 or dst host 192.168.3.115)' -c 2000 -s0 -w /tmp/savefile-fifo
 </pre>

← Older revision		Revision as of 04:40, 1 December 2014
Line 25:		Line 25:

	== Results ==		== Results ==
		+
		+	The test setup was as follows:
		+	* Intel i5-3450, running Debian Wheezy, set to send ArtNet traffic on universe 1
		+	* Raspberry Pi, configured to receive ArtNet traffic and convert to sACN traffic on universe 1.
		+	* Both hosts connect to a netgear 5 port switch. The Pi was connected at 100Mbps, the Debian host at 1Gbps.
		+
		+	On the Debian host, the traffic was generated with:
		+
		+	<pre>
		+	$ i=0; while :; do echo "$i"; i=$((i+1)); if [ $i == 256 ] ; then i=0; fi ; sleep 0.1; done \| ola_streaming_client -u 1
		+	</pre>
		+
		+	And the packets captured with:
		+
		+	<pre>
		+	tcpdump port 5568 or port 6454 -n -s0 -w /tmp/savefile
		+	</pre>
		+
		+	On the Pi, the control was with no scheduling options. Then the fifo scheduler was enabled with
		+
		+	<pre>
		+	$ olad --scheduler-policy fifo --scheduler-priority 10
		+	</pre>

@@ Line 1: / Line 1: @@
 Since 0.9.4, OLA provides options to control the thread scheduling algorithms, when supported by the host kernel.
-Besides the default scheduling policy (SCHED_OTHER), OLA supports SCHED_RR and SCHED_FIFO. You can read about the differences on [http://www.linuxjournal.com/article/3910 | The Linux Journal].
+Besides the default scheduling policy (SCHED_OTHER), OLA supports SCHED_RR and SCHED_FIFO. You can read about the differences on [http://www.linuxjournal.com/article/3910 The Linux Journal].
 == Configuration ==