Can't get Multithreaded Nashorn uses to Scale
Jesus Luzon
jluzon at riotgames.com
Fri Dec 9 00:54:57 UTC 2016
It doesn't seem like changing to a string fixed it. We dug deeper and found
a very interesting issue though. After running the invocable a few thousand
times, when we put a breakpoint in ArrayData.computerIteratorKeys() and we
found that it would stop at an execution from SparseArrayData with an
underlying array of size 923392, stored length of 461672 (this is one of
our IDs, the largest 6 digit one). Because of this we tried a run changing
all the IDs in the array we pass in to number from 1-38 and this thing went
blazing fast. When putting a break point in the same place, this array in
SpareArrayData was now of size 39. We then changed an id in the array we
take in to be 1337 and the size of the array in the SparseArrayData was
1338.
I don't understand why or how to prevent but it's using this ID as an index
for SpareArrayData underlying array. If someone can help me find a
workaround for this I would be extremely grateful.
On Thu, Dec 8, 2016 at 3:52 AM, Hannes Wallnöfer <
hannes.wallnoefer at oracle.com> wrote:
> Yes, this is very likely to be the cause of the problem. However, I do
> think we should be able to handle sparse array data better, so it’s quite
> possible that you exposed some weak spot in our implementation. I’ll have a
> look into what’s going on.
>
> Hannes
>
>
> > Am 08.12.2016 um 00:40 schrieb Jesus Luzon <jluzon at riotgames.com>:
> >
> > Looks like the memory leak is due to the way we wrote our code and how
> javascript works. I was expecting the line response[summoner.id] =
> summoner; to build a map but it turns out that if you use a number as the
> "key", javscript automatically fills the indexes in the middle with null
> (Undefined type?). When these IDs are very large, it is creating huge
> arrays that take longer to garbage collect than the code executing. I am
> about to start testing this on our end to make sure we see the improvements
> we expect.
> >
> > Does this idea seem like it is reasonably possible?
> >
> > On Wed, Dec 7, 2016 at 11:23 AM, Jesus Luzon <jluzon at riotgames.com>
> wrote:
> > Yes, it's an array of objects which I'll paste. And yes, I'm just
> calling invokeFunction from many many different threads. I'm also going to
> go back and take a look at all the heap dumps we have to re-confirm what I
> mentioned.
> >
> > "[\n" +
> > " {\n" +
> > " \"id\": 6011511,\n" +
> > " \"accountId\": 203192481,\n" +
> > " \"name\": \"Adam Pro Qardaş\",\n" +
> > " \"profileIconId\": 25,\n" +
> > " \"level\": 5,\n" +
> > " \"expPoints\": 83,\n" +
> > " \"infPoints\": 1475,\n" +
> > " \"revisionDate\": 1406631727000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 2810674,\n" +
> > " \"accountId\": 200706913,\n" +
> > " \"name\": \"ABZ Devrim\",\n" +
> > " \"profileIconId\": 663,\n" +
> > " \"level\": 13,\n" +
> > " \"expPoints\": 982,\n" +
> > " \"infPoints\": 10472,\n" +
> > " \"revisionDate\": 1450791227000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5411195,\n" +
> > " \"accountId\": 202647689,\n" +
> > " \"name\": \"Ace HypcronN\",\n" +
> > " \"profileIconId\": 911,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 73,\n" +
> > " \"infPoints\": 182445,\n" +
> > " \"revisionDate\": 1480781650000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1363020,\n" +
> > " \"accountId\": 1357837,\n" +
> > " \"name\": \"AdanaLee\",\n" +
> > " \"profileIconId\": 502,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 125,\n" +
> > " \"infPoints\": 719299,\n" +
> > " \"revisionDate\": 1480530778000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 8261198,\n" +
> > " \"accountId\": 205027096,\n" +
> > " \"name\": \"Achilehuz\",\n" +
> > " \"profileIconId\": 1381,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 10,\n" +
> > " \"infPoints\": 158603,\n" +
> > " \"revisionDate\": 1480770307000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 12685857,\n" +
> > " \"accountId\": 207591166,\n" +
> > " \"name\": \"acımasızpicc\",\n" +
> > " \"profileIconId\": 9,\n" +
> > " \"level\": 21,\n" +
> > " \"expPoints\": 840,\n" +
> > " \"infPoints\": 16659,\n" +
> > " \"revisionDate\": 1480515325000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 10860127,\n" +
> > " \"accountId\": 206507727,\n" +
> > " \"name\": \"AAngelFlyy\",\n" +
> > " \"profileIconId\": 1395,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 10,\n" +
> > " \"infPoints\": 73111,\n" +
> > " \"revisionDate\": 1480787870000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 3292376,\n" +
> > " \"accountId\": 201048714,\n" +
> > " \"name\": \"ACAB1907\",\n" +
> > " \"profileIconId\": 20,\n" +
> > " \"level\": 6,\n" +
> > " \"expPoints\": 305,\n" +
> > " \"infPoints\": 2107,\n" +
> > " \"revisionDate\": 1402448089000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 461671,\n" +
> > " \"accountId\": 446571,\n" +
> > " \"name\": \"Acta Est Fabulâ\",\n" +
> > " \"profileIconId\": 1435,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 47,\n" +
> > " \"infPoints\": 644672,\n" +
> > " \"revisionDate\": 1480626505000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 394183,\n" +
> > " \"accountId\": 379083,\n" +
> > " \"name\": \"acekse4\",\n" +
> > " \"profileIconId\": 27,\n" +
> > " \"level\": 5,\n" +
> > " \"expPoints\": 223,\n" +
> > " \"infPoints\": 908,\n" +
> > " \"revisionDate\": 1348116544000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5941247,\n" +
> > " \"accountId\": 203106300,\n" +
> > " \"name\": \"abdul7878\",\n" +
> > " \"profileIconId\": 26,\n" +
> > " \"level\": 3,\n" +
> > " \"expPoints\": 10,\n" +
> > " \"infPoints\": 401,\n" +
> > " \"revisionDate\": 1406029148000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 2467446,\n" +
> > " \"accountId\": 200459837,\n" +
> > " \"name\": \"ActionC\",\n" +
> > " \"profileIconId\": 986,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 74,\n" +
> > " \"infPoints\": 401367,\n" +
> > " \"revisionDate\": 1480808608000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 9402979,\n" +
> > " \"accountId\": 205698832,\n" +
> > " \"name\": \"Ablenia \",\n" +
> > " \"profileIconId\": 1129,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 19,\n" +
> > " \"infPoints\": 163518,\n" +
> > " \"revisionDate\": 1480687603000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 13187505,\n" +
> > " \"accountId\": 207898213,\n" +
> > " \"name\": \"aDaMiYiM\",\n" +
> > " \"profileIconId\": 1301,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 116,\n" +
> > " \"infPoints\": 45214,\n" +
> > " \"revisionDate\": 1480793258000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 4141059,\n" +
> > " \"accountId\": 201688290,\n" +
> > " \"name\": \"AbiminÇarı\",\n" +
> > " \"profileIconId\": 898,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 152,\n" +
> > " \"infPoints\": 752477,\n" +
> > " \"revisionDate\": 1480635961000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5702134,\n" +
> > " \"accountId\": 202899395,\n" +
> > " \"name\": \"Above the Clouds\",\n" +
> > " \"profileIconId\": 684,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 110,\n" +
> > " \"infPoints\": 288096,\n" +
> > " \"revisionDate\": 1471011372000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5810740,\n" +
> > " \"accountId\": 202985228,\n" +
> > " \"name\": \"aBimm\",\n" +
> > " \"profileIconId\": 11,\n" +
> > " \"level\": 13,\n" +
> > " \"expPoints\": 1180,\n" +
> > " \"infPoints\": 10736,\n" +
> > " \"revisionDate\": 1409832684000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5817751,\n" +
> > " \"accountId\": 203050678,\n" +
> > " \"name\": \"AD Glorıam\",\n" +
> > " \"profileIconId\": 982,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 111,\n" +
> > " \"infPoints\": 304658,\n" +
> > " \"revisionDate\": 1480795250000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 9851802,\n" +
> > " \"accountId\": 206011054,\n" +
> > " \"name\": \"AdarAllame\",\n" +
> > " \"profileIconId\": 911,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 48,\n" +
> > " \"infPoints\": 73763,\n" +
> > " \"revisionDate\": 1479422812000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 12735622,\n" +
> > " \"accountId\": 207587019,\n" +
> > " \"name\": \"absinthe666\",\n" +
> > " \"profileIconId\": 903,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 83,\n" +
> > " \"infPoints\": 40302,\n" +
> > " \"revisionDate\": 1480782923000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 6371389,\n" +
> > " \"accountId\": 203416952,\n" +
> > " \"name\": \"adamsatıcı\",\n" +
> > " \"profileIconId\": 3,\n" +
> > " \"level\": 4,\n" +
> > " \"expPoints\": 17,\n" +
> > " \"infPoints\": 685,\n" +
> > " \"revisionDate\": 1409320171000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 7828139,\n" +
> > " \"accountId\": 204927980,\n" +
> > " \"name\": \"AbsoluteForce\",\n" +
> > " \"profileIconId\": 950,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 139,\n" +
> > " \"infPoints\": 208789,\n" +
> > " \"revisionDate\": 1480804396000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1373229,\n" +
> > " \"accountId\": 1358441,\n" +
> > " \"name\": \"AbsoluteDeath\",\n" +
> > " \"profileIconId\": 7,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 34,\n" +
> > " \"infPoints\": 223655,\n" +
> > " \"revisionDate\": 1471867646000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 7694972,\n" +
> > " \"accountId\": 204803668,\n" +
> > " \"name\": \"ac pnp\",\n" +
> > " \"profileIconId\": 937,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 161,\n" +
> > " \"infPoints\": 249681,\n" +
> > " \"revisionDate\": 1480801507000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1373524,\n" +
> > " \"accountId\": 1350474,\n" +
> > " \"name\": \"Abdüü\",\n" +
> > " \"profileIconId\": 1301,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 103,\n" +
> > " \"infPoints\": 286803,\n" +
> > " \"revisionDate\": 1476621827000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1650227,\n" +
> > " \"accountId\": 200000503,\n" +
> > " \"name\": \"AD Ambrosia\",\n" +
> > " \"profileIconId\": 1152,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 139,\n" +
> > " \"infPoints\": 156333,\n" +
> > " \"revisionDate\": 1480805320000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 8331358,\n" +
> > " \"accountId\": 205073925,\n" +
> > " \"name\": \"acarmanyust2\",\n" +
> > " \"profileIconId\": 0,\n" +
> > " \"level\": 2,\n" +
> > " \"expPoints\": 43,\n" +
> > " \"infPoints\": 318,\n" +
> > " \"revisionDate\": 1423915139000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1862106,\n" +
> > " \"accountId\": 200139838,\n" +
> > " \"name\": \"aboU\",\n" +
> > " \"profileIconId\": 1155,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 0,\n" +
> > " \"infPoints\": 412616,\n" +
> > " \"revisionDate\": 1480771055000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 2362628,\n" +
> > " \"accountId\": 685649,\n" +
> > " \"name\": \"AcıFısTık\",\n" +
> > " \"profileIconId\": 1074,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 48,\n" +
> > " \"infPoints\": 233882,\n" +
> > " \"revisionDate\": 1480786233000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 4323909,\n" +
> > " \"accountId\": 201917672,\n" +
> > " \"name\": \"Addrenalin\",\n" +
> > " \"profileIconId\": 603,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 55,\n" +
> > " \"infPoints\": 220605,\n" +
> > " \"revisionDate\": 1432647338000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 377206,\n" +
> > " \"accountId\": 362106,\n" +
> > " \"name\": \"Aburame Shino\",\n" +
> > " \"profileIconId\": 844,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 84,\n" +
> > " \"infPoints\": 354087,\n" +
> > " \"revisionDate\": 1477666556000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5377433,\n" +
> > " \"accountId\": 202697921,\n" +
> > " \"name\": \"AcEcolton35\",\n" +
> > " \"profileIconId\": 984,\n" +
> > " \"level\": 25,\n" +
> > " \"expPoints\": 751,\n" +
> > " \"infPoints\": 30061,\n" +
> > " \"revisionDate\": 1475503024000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 2381404,\n" +
> > " \"accountId\": 200333680,\n" +
> > " \"name\": \"adafakaaa\",\n" +
> > " \"profileIconId\": 663,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 8,\n" +
> > " \"infPoints\": 534204,\n" +
> > " \"revisionDate\": 1480719827000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 1281203,\n" +
> > " \"accountId\": 1259342,\n" +
> > " \"name\": \"AC Klondike\",\n" +
> > " \"profileIconId\": 898,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 27,\n" +
> > " \"infPoints\": 191429,\n" +
> > " \"revisionDate\": 1480294973000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 13161471,\n" +
> > " \"accountId\": 207847181,\n" +
> > " \"name\": \"adar21\",\n" +
> > " \"profileIconId\": 26,\n" +
> > " \"level\": 10,\n" +
> > " \"expPoints\": 143,\n" +
> > " \"infPoints\": 3558,\n" +
> > " \"revisionDate\": 1476529855000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 5841915,\n" +
> > " \"accountId\": 202998794,\n" +
> > " \"name\": \"Achilles29\",\n" +
> > " \"profileIconId\": 666,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 41,\n" +
> > " \"infPoints\": 219714,\n" +
> > " \"revisionDate\": 1480777744000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 2853062,\n" +
> > " \"accountId\": 200726707,\n" +
> > " \"name\": \"AbIanStarBebegim\",\n" +
> > " \"profileIconId\": 909,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 64,\n" +
> > " \"infPoints\": 297580,\n" +
> > " \"revisionDate\": 1480556859000\n" +
> > " },\n" +
> > " {\n" +
> > " \"id\": 8323114,\n" +
> > " \"accountId\": 205093515,\n" +
> > " \"name\": \"Absuruk\",\n" +
> > " \"profileIconId\": 6,\n" +
> > " \"level\": 30,\n" +
> > " \"expPoints\": 21,\n" +
> > " \"infPoints\": 121086,\n" +
> > " \"revisionDate\": 1480820801000\n" +
> > " }\n" +
> > "]"
> >
> > On Wed, Dec 7, 2016 at 7:01 AM, Hannes Wallnöfer <
> hannes.wallnoefer at oracle.com> wrote:
> > Hi Jesus,
> >
> > I’m trying to reproduce the problem, and just want to make sure I get
> the missing pieces right.
> >
> > You already showed us how you’re setting up the engine and the JS code
> you’re running. I assume the JSON code you’re parsing is a simple array of
> objects? And you’re just calling Invocable.invokeFunction on the
> ScriptEngine from multiple threads in parallel, right?
> >
> > Thanks,
> > Hannes
> >
> >
> > > Am 07.12.2016 um 00:03 schrieb Jesus Luzon <jluzon at riotgames.com>:
> > >
> > > When we share one invocable across many threads and run invokeFunction
> it
> > > happens, such as this:
> > >
> > > ExecutorService executor = Executors.newFixedThreadPool(50);
> > >>
> > >> Invocable invocable = generateInvocable(script);
> > >>
> > >> AtomicLong count = new AtomicLong();
> > >>
> > >> for (int i = 0; i < 50; i++) {
> > >>
> > >> executor.submit(new Runnable() {
> > >>
> > >> @Override
> > >>
> > >> public void run() {
> > >>
> > >> try {
> > >>
> > >> while(true) {
> > >>
> > >> invocable.invokeFunction("transform",
> > >>> something);
> > >>
> > >> count.incrementAndGet();
> > >>
> > >> }
> > >>
> > >> } catch (NoSuchMethodException |
> ScriptException
> > >>> e) {
> > >>
> > >> e.printStackTrace();
> > >>
> > >> }
> > >>
> > >> }
> > >>
> > >> });
> > >>
> > >> }
> > >>
> > >>
> > >
> > >
> > > On Tue, Dec 6, 2016 at 2:59 PM, Jim Laskey (Oracle) <
> james.laskey at oracle.com
> > >> wrote:
> > >
> > >> Intersting. The example you posted demonstrates this behaviour? If
> so
> > >> I’ll file a bug and dig in. It sounds like an object is being reused
> > >> across invocations and accumulating changes to the property map.
> > >>
> > >> — Jim
> > >>
> > >>
> > >> On Dec 6, 2016, at 5:12 PM, Jesus Luzon <jluzon at riotgames.com> wrote:
> > >>
> > >> With more threads you are impacting the same 8 cores, so it will
> taper off
> > >>> after 8 threads. If it’s a 2x4 core machine then I can see 4 being a
> > >>> threshold depending on System performance. Transport: I meant if
> you were
> > >>> using sockets to provide the script.
> > >>
> > >> This makes sense. This one's on me then.
> > >>
> > >>
> > >>> So you are using the same invocable instance for all threads? If so,
> > >>> then you are probably good to go. As far as leaks are concerned,
> not sure
> > >>> how you would get leaks from Nashorn. The JSON object is written in
> Java,
> > >>> and little JavaScript involved.
> > >>
> > >>
> > >>
> > >>> In your example, pull up Invocable invocable =
> generateInvocable(script);
> > >>> out of the loop and use the same invocable for all threads.
> > >>
> > >>
> > >> We were using one invocable across all threads and we were getting
> > >> slowdowns on execution, high CPU Usage and memory leaks that led to
> > >> OutOfMemory errors. I could trace the leak to
> > >>
> > >> jdk.nashorn.internal.objects.Global -> *objectSpill* Object[8] ->
> > >> jdk.nashorn.internal.scripts.JO4 -> *arrayData*
> > >> jdk.nashorn.internal.runtime.arrays.SparseArraysData -> *underlying*
> > >> jdk.nashorn.internal.runtime.arrays.DeletedArrayFilter
> > >>
> > >> which just keeps growing forever.
> > >>
> > >> On Tue, Dec 6, 2016 at 6:30 AM, Jim Laskey (Oracle) <
> > >> james.laskey at oracle.com> wrote:
> > >>
> > >>>
> > >>> On Dec 6, 2016, at 9:56 AM, Jesus Luzon <jluzon at riotgames.com>
> wrote:
> > >>>
> > >>> The cost of creating a new engine is significant. So share an engine
> > >>>> across threads but use *eval
> > >>>> <https://docs.oracle.com/javase/7/docs/api/javax/
> script/ScriptEngine.html#eval(java.lang.String,%20javax.
> script.ScriptContext)>*
> > >>>> (String
> > >>>> <https://docs.oracle.com/javase/7/docs/api/java/lang/String.html>
> > >>>> script, ScriptContext
> > >>>> <https://docs.oracle.com/javase/7/docs/api/javax/
> script/ScriptContext.html>
> > >>>> context) instead, separate context per execution. If your
> JavaScript
> > >>>> code does not modify globals you can get away with using the same
> engine,
> > >>>> same compiled script on each thread.
> > >>>
> > >>>
> > >>> I guess there's a few things here I don't understand. One thing I'm
> > >>> trying to do is sharing a CompiledScript (which is why I'm using
> > >>> invocable). Also, what exactly does modify globals mean? All our
> filters do
> > >>> the same thing, make a function that takes a JSON String, turns it
> into a
> > >>> JSON, modifies it and then stringifies it back. No state is changed
> of
> > >>> anything else but there are temporary vars created inside the scope
> of the
> > >>> function. When we run this multithreaded, running invokeFunction
> slows down
> > >>> significantly and we get crazy memory leaks.
> > >>>
> > >>>
> > >>> So you are using the same invocable instance for all threads? If so,
> > >>> then you are probably good to go. As far as leaks are concerned,
> not sure
> > >>> how you would get leaks from Nashorn. The JSON object is written in
> Java,
> > >>> and little JavaScript involved.
> > >>>
> > >>>
> > >>> Of course there are many factors involved n performance. How many
> cores
> > >>>> do you have on the test machine? How much memory in the process?
> What
> > >>>> transport are you using between threads? That sort of thing.
> Other than
> > >>>> constructing then engine and context Nashorn performance should
> scale.
> > >>>
> > >>> I'm using an 8 core machine to test with 2.5Gs of RAM allocated to
> the
> > >>> process. Not sure what transports between threads means, but this is
> the
> > >>> code I'm benchmarking with. Increasing the number of threads
> actually makes
> > >>> it go faster until about 4 threads, then adding more threads takes
> the same
> > >>> amount to get to 1000 and and after a certain point it is just
> slower to
> > >>> get to 1000 counts. Some of our filters need to be able to run over
> 1000
> > >>> times a second (across all threads) and the fastest time I could
> actually
> > >>> get with this was about 2.4 seconds for a 1000 counts.
> > >>>
> > >>>> ExecutorService executor = Executors.newFixedThreadPool(50);
> > >>>>
> > >>>> AtomicLong count = new AtomicLong();
> > >>>>
> > >>>> for (int i = 0; i < 50; i++) {
> > >>>>
> > >>>> executor.submit(new Runnable() {
> > >>>>
> > >>>> @Override
> > >>>>
> > >>>> public void run() {
> > >>>>
> > >>>>
> > >>>>> try {
> > >>>>
> > >>>> Invocable invocable =
> > >>>>> generateInvocable(script);
> > >>>>
> > >>>> while(true) {
> > >>>>
> > >>>> invocable.invokeFunction("
> transform",
> > >>>>> something);
> > >>>>
> > >>>> count.incrementAndGet();
> > >>>>
> > >>>> }
> > >>>>
> > >>>> } catch (NoSuchMethodException |
> ScriptException
> > >>>>> e) {
> > >>>>
> > >>>> e.printStackTrace();
> > >>>>
> > >>>> }
> > >>>>
> > >>>> }
> > >>>>
> > >>>> });
> > >>>>
> > >>>> }
> > >>>>
> > >>>> long lastTimestamp = System.currentTimeMillis();
> > >>>>
> > >>>> while(true) {
> > >>>>
> > >>>>
> > >>>>> if (count.get() > 1000) {
> > >>>>
> > >>>> count.getAndAdd(-1000);
> > >>>>
> > >>>> System.out.println((System.currentTimeMillis() -
> > >>>>> lastTimestamp)/1000.0);
> > >>>>
> > >>>> lastTimestamp = System.currentTimeMillis();
> > >>>>
> > >>>> }
> > >>>>
> > >>>> }
> > >>>>
> > >>>>
> > >>> With more threads you are impacting the same 8 cores, so it will
> taper
> > >>> off after 8 threads. If it’s a 2x4 core machine then I can see 4
> being a
> > >>> threshold depending on System performance. Transport: I meant if
> you were
> > >>> using sockets to provide the script.
> > >>>
> > >>> In your example, pull up Invocable invocable =
> generateInvocable(script);
> > >>> out of the loop and use the same invocable for all threads.
> > >>>
> > >>> - Jim
> > >>>
> > >>>
> > >>>
> > >>>
> > >>> On Tue, Dec 6, 2016 at 5:31 AM, Jim Laskey (Oracle) <
> > >>> james.laskey at oracle.com> wrote:
> > >>>
> > >>>>
> > >>>> On Dec 6, 2016, at 9:19 AM, Jesus Luzon <jluzon at riotgames.com>
> wrote:
> > >>>>
> > >>>> Hey Jim,
> > >>>>
> > >>>> I looked at it and I will look into loadWithNewGlobal to see what
> > >>>> exactly it does since it could be relevant. As for the rest, for my
> use
> > >>>> case having threads in the JS would not help. We're using Nashorn
> to build
> > >>>> JSON filters in a Dynamic Proxy Service and need any of the threads
> > >>>> processing a request to be able to execute the script to filter.
> > >>>>
> > >>>>
> > >>>> The cost of creating a new engine is significant. So share an
> engine
> > >>>> across threads but use *eval
> > >>>> <https://docs.oracle.com/javase/7/docs/api/javax/
> script/ScriptEngine.html#eval(java.lang.String,%20javax.
> script.ScriptContext)>*
> > >>>> (String
> > >>>> <https://docs.oracle.com/javase/7/docs/api/java/lang/String.html>
> > >>>> script, ScriptContext
> > >>>> <https://docs.oracle.com/javase/7/docs/api/javax/
> script/ScriptContext.html>
> > >>>> context) instead, separate context per execution. If your
> JavaScript
> > >>>> code does not modify globals you can get away with using the same
> engine,
> > >>>> same compiled script on each thread.
> > >>>>
> > >>>>
> > >>>> Also, when you say a new engine per threads is the worst case what
> > >>>> exactly do you mean? I would expect an initial cost of compiling
> the script
> > >>>> on each thread and then each engine should be able to do its own
> thing, but
> > >>>> what I'm seeing is that when running with more than 10 threads all
> my
> > >>>> engines get slow at executing code, even though they are all
> completely
> > >>>> separate from each other.
> > >>>>
> > >>>>
> > >>>> Of course there are many factors involved n performance. How many
> cores
> > >>>> do you have on the test machine? How much memory in the process?
> What
> > >>>> transport are you using between threads? That sort of thing.
> Other than
> > >>>> constructing then engine and context Nashorn performance should
> scale.
> > >>>>
> > >>>>
> > >>>> On Tue, Dec 6, 2016 at 5:07 AM, Jim Laskey (Oracle) <
> > >>>> james.laskey at oracle.com> wrote:
> > >>>>
> > >>>>> Jesus,
> > >>>>>
> > >>>>> Probably the most informative information is in this blog.
> > >>>>>
> > >>>>> https://blogs.oracle.com/nashorn/entry/nashorn_multi_
> threading_and_mt
> > >>>>>
> > >>>>> This example uses Executors but threads would work as well.
> > >>>>>
> > >>>>> I did a talk that looked at different methods to max out
> multithreading
> > >>>>> performance. A new engine per thread is the worst case. A new
> context per
> > >>>>> thread does much better. A new global per thread is the best while
> > >>>>> remaining thread safe.
> > >>>>>
> > >>>>> Cheers,
> > >>>>>
> > >>>>> — Jim
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>> On Dec 6, 2016, at 8:45 AM, Jesus Luzon <jluzon at riotgames.com>
> wrote:
> > >>>>>
> > >>>>> Hey folks,
> > >>>>>
> > >>>>> I've tried many different ways of using Nashorn multithreaded
> based on
> > >>>>> what
> > >>>>> I've found on the internet and I still can't get a single one to
> scale.
> > >>>>> Even the most naive method of making many script engines with my
> script
> > >>>>> tends to bottleneck itself when I have more than 10 threads
> invoking
> > >>>>> functions.
> > >>>>>
> > >>>>> I'm using the following code to compile my script and
> > >>>>> invocable.invokeFunction("transform", input) to execute:
> > >>>>>
> > >>>>> static Invocable generateInvocable(String script) throws
> > >>>>> ScriptException {
> > >>>>> ScriptEngineManager manager = new ScriptEngineManager();
> > >>>>> ScriptEngine engine =
> > >>>>> manager.getEngineByName(JAVASCRIPT_ENGINE_NAME);
> > >>>>> Compilable compilable = (Compilable) engine;
> > >>>>> final CompiledScript compiled = compilable.compile(script);
> > >>>>> compiled.eval();
> > >>>>> return (Invocable) engine;
> > >>>>> }
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>> The script I'm compiling is:
> > >>>>>
> > >>>>> String script = "function transform(input) {" +
> > >>>>> "var result = JSON.parse(input);" +
> > >>>>> "response = {};\n" +
> > >>>>> "for (var i = 0; i < result.length; i++) {\n" +
> > >>>>> " var summoner = {};\n" +
> > >>>>> " summoner.id = result[i].id;\n" +
> > >>>>> " summoner.name = result[i].name;\n" +
> > >>>>> " summoner.profileIconId =
> > >>>>> result[i].profileIconId;\n" +
> > >>>>> " summoner.revisionDate =
> result[i].revisionDate;\n" +
> > >>>>> " summoner.summonerLevel = result[i].level;\n" +
> > >>>>> " response[summoner.id] = summoner;\n" +
> > >>>>> "}\n" +
> > >>>>> "result = response;" +
> > >>>>> "return JSON.stringify(result);" +
> > >>>>> "};";
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>> I've also tried other more scaleable ways to work with scripts
> > >>>>> concurrently, but given that this is the most naive method where
> > >>>>> everything
> > >>>>> is brand new and I still get slowness calling them concurrently I
> fear
> > >>>>> that
> > >>>>> maybe I'm overlooking something extremely basic on my code.
> > >>>>>
> > >>>>> Thanks.
> > >>>>> -Jesus Luzon
> > >>>>>
> > >>>>>
> > >>>>
> > >>>
> > >>>
> > >>
> > >>
> >
> >
> >
>
>
More information about the nashorn-dev
mailing list