read Environment.ProcessorCount once into a static

[0x53A]

fixes #2713

[0x53A]

ref fsprojects/FAKE#1571 for the build error ...

[Danthar]

Not sure why test are failing. I restarted the tests,
The change is harmless from a functional perspective.

[Aaronontheweb]

Before

Akka.Remote.Tests.Performance.Transports.HeliosRemoteMessagingThroughputSpec+OneWay

Measures the throughput of Akka.Remote over a particular transport using one-way messaging
5/30/2017 10:58:53 PM

System Info

NBench=NBench, Version=1.0.1.0, Culture=neutral, PublicKeyToken=null
OS=Microsoft Windows NT 6.2.9200.0
ProcessorCount=2
CLR=4.0.30319.42000,IsMono=False,MaxGcGeneration=2

NBench Settings

RunMode=Iterations, TestMode=Measurement
NumberOfIterations=3, MaximumRunTime=00:00:01
Concurrent=True
Tracing=True

Data

---

Totals

Metric	Units	Max	Average	Min	StdDev
TotalCollections [Gen0]	collections	137.00	136.67	136.00	0.58
TotalCollections [Gen1]	collections	4.00	2.33	1.00	1.53
TotalCollections [Gen2]	collections	0.00	0.00	0.00	0.00
[Counter] RemoteMessageReceived	operations	50,000.00	50,000.00	50,000.00	0.00

Per-second Totals

Metric	Units / s	Max / s	Average / s	Min / s	StdDev / s
TotalCollections [Gen0]	collections	25.58	25.13	24.87	0.40
TotalCollections [Gen1]	collections	0.73	0.43	0.19	0.27
TotalCollections [Gen2]	collections	0.00	0.00	0.00	0.00
[Counter] RemoteMessageReceived	operations	9,336.66	9,192.27	9,075.48	132.76

Raw Data

TotalCollections [Gen0]

Run #	collections	collections / s	ns / collections
1	137.00	24.87	40,214,253.28
2	137.00	25.58	39,089,308.76
3	136.00	24.93	40,115,722.79

TotalCollections [Gen1]

Run #	collections	collections / s	ns / collections
1	4.00	0.73	1,377,338,175.00
2	1.00	0.19	5,355,235,300.00
3	2.00	0.37	2,727,869,150.00

TotalCollections [Gen2]

Run #	collections	collections / s	ns / collections
1	0.00	0.00	5,509,352,700.00
2	0.00	0.00	5,355,235,300.00
3	0.00	0.00	5,455,738,300.00

[Counter] RemoteMessageReceived

Run #	operations	operations / s	ns / operations
1	50,000.00	9,075.48	110,187.05
2	50,000.00	9,336.66	107,104.71
3	50,000.00	9,164.66	109,114.77

[Aaronontheweb]

After

Akka.Remote.Tests.Performance.Transports.HeliosRemoteMessagingThroughputSpec+OneWay

Measures the throughput of Akka.Remote over a particular transport using one-way messaging
5/31/2017 8:51:58 AM

System Info

NBench=NBench, Version=0.3.4.0, Culture=neutral, PublicKeyToken=null
OS=Microsoft Windows NT 6.2.9200.0
ProcessorCount=2
CLR=4.0.30319.42000,IsMono=False,MaxGcGeneration=2
WorkerThreads=32767, IOThreads=2

NBench Settings

RunMode=Iterations, TestMode=Measurement
NumberOfIterations=3, MaximumRunTime=00:00:01
Concurrent=True
Tracing=True

Data

---

Totals

Metric	Units	Max	Average	Min	StdDev
TotalCollections [Gen0]	collections	213.00	204.00	197.00	8.19
TotalCollections [Gen1]	collections	28.00	23.67	20.00	4.04
TotalCollections [Gen2]	collections	1.00	0.33	0.00	0.58
[Counter] RemoteMessageReceived	operations	50,000.00	50,000.00	50,000.00	0.00

Per-second Totals

Metric	Units / s	Max / s	Average / s	Min / s	StdDev / s
TotalCollections [Gen0]	collections	63.13	62.38	61.12	1.10
TotalCollections [Gen1]	collections	8.97	7.27	6.05	1.52
TotalCollections [Gen2]	collections	0.32	0.11	0.00	0.19
[Counter] RemoteMessageReceived	operations	16,022.57	15,303.90	14,761.50	648.75

Raw Data

TotalCollections [Gen0]

Run #	collections	collections / s	ns / collections
1	213.00	62.88	15,902,298.59
2	197.00	63.13	15,840,596.45
3	202.00	61.12	16,362,440.59

TotalCollections [Gen1]

Run #	collections	collections / s	ns / collections
1	23.00	6.79	147,269,113.04
2	28.00	8.97	111,449,910.71
3	20.00	6.05	165,260,650.00

TotalCollections [Gen2]

Run #	collections	collections / s	ns / collections
1	0.00	0.00	3,387,189,600.00
2	1.00	0.32	3,120,597,500.00
3	0.00	0.00	3,305,213,000.00

[Counter] RemoteMessageReceived

Run #	operations	operations / s	ns / operations
1	50,000.00	14,761.50	67,743.79
2	50,000.00	16,022.57	62,411.95
3	50,000.00	15,127.62	66,104.26

[Aaronontheweb]

Specs look good from the POV of Akka.Remote, but the current dispatcher throughput specs that measure the dispatcher implementations themselves don't perform enough work for a measurement to be reliable. About to submit a PR that bumps it up considerably.

[Aaronontheweb]

@Danthar some sort of issue with the Mono compiler... No idea why that's having an issue, given that it appears to be having trouble with FAKE.

cc @heynickc did we change the Mono docker container at all today?

[Aaronontheweb]

@0x53A looks like the build agent was properly cleaning itself out on checkout for Mono builds, hence why we had this issue. Just updated those settings and am re-running it now.

[Aaronontheweb]

No idea where there's a persistent Mono compilation failure here, given that we're doing clean checkouts now...

[0x53A]

It looks like it fails when it Writes the cache, not when Reading it.

If you are doing a clean build anyway, I would just disable the cache.

[Aaronontheweb]

@0x53A killing off the current TC build agents and relaunching them since the same issue re-appeared

[0x53A]

My guess that it's related to the cache was wrong, it can't even compile build.fsx:
http://petabridge-ci.cloudapp.net/viewLog.html?buildId=27587&buildTypeId=AkkaNet_LinuxMonoBuild&tab=buildLog#_state=684,695,703,30,35,32,354,395&focus=618

This looks similar to fsharp/fsharp#685, which was fixed in fsharp/fsharp#687.

---

This would also explain why your pr didn't fail - mine is based on master, which had an old FAKE, yours was against 1.3, which has a newer FAKE.

[Aaronontheweb]

@0x53A you were targeting the master branch, which explains a bunch of these issues - I've changed it to point to the v1.3 branch instead :p

Could you revert your changes to the build scripts?

[Aaronontheweb]

@0x53A actually, I'd probably just rebase the entire thing on v1.3 and remove those two commits

[0x53A]

@Aaronontheweb done - let's see if the CI likes me

[0x53A]

@Aaronontheweb build looks broken:

[Aaronontheweb]

@0x53A we're having issues with other parts of our test suite unrelated to your changes (timeouts;) no worries, we'll b dealing with that shortly.