I have two edge collections: edges_temp with 17 millions edges and myedges with 22 millions edges
I'm doing the following:
FOR edge IN edges_temp
UPSERT { _key:edges_temp._key }
INSERT { _from: edge._from, _to: edge._to, coccurrence: edge.coccurrence, _key:edges_temp._key }
UPDATE { coccurrence: OLD.coccurrence + edge.coccurrence }
IN myedges
And getting the error
Query: AQL: out of memory (exception location:
C:\b\workspace\RELEASE__BuildPackages\lib\V8\v8-vpack.cpp:125)
stacktrace of offending AQL function: ArangoError: out of memory
(exception location:
C:\b\workspace\RELEASE__BuildPackages\lib\V8\v8-vpack.cpp:125) at
Object.GET_DOCUMENTS (c:\Program Files\ArangoDB3
3.3.4\usr\share\arangodb3\js\server\modules\#arangodb\aql.js:1182:39) at Function. (--script--:1:183) (exception location:
C:\b\workspace\RELEASE__BuildPackages\arangod\Aql\V8Executor.cpp:333)
(while executing) (exception location:
C:\b\workspace\RELEASE__BuildPackages\arangod\RestHandler\RestCursorHandler.cpp:134)
I understand that they are quite a big collections which are loaded in RAM, but at the moment I get the error I have ~4GB free RAM.
What else could cause that exception?
UPD:
OS: win10
ArangoDB v.3.3.4
All settings were by default, storage engine: auto, single machine.
Related
I am trying to load test a Kafka instance on one of our servers.
Here is the code that does it using NBomber:
public static void Run()
{
var testScenario = NBomber.CSharp.Step.Create("testScenario",
async context =>
{
try
{
// The testData is a string variable that reads contents from a text file in the Init method.
var kafkaObject = new KafkaObject { Topic = TestTopic, Message =testData };
SampleKafkaFlow sampleKafkaFlow = new SampleKafkaFlow();
var response = await sampleKafkaFlow.SendMessageToKafka(kafkaObject);
return Response.Ok();
}
catch (Exception ex)
{
return Response.Fail(ex.Message);
}
});
var scenario = ScenarioBuilder.CreateScenario("scenario", testScenario)
.WithoutWarmUp()
.WithInit(Init)
.WithLoadSimulations(new[]
{
Simulation.InjectPerSec(rate: 100, during: TimeSpan.FromMinutes(3))
});
NBomber.CSharp.NBomberRunner
.RegisterScenarios(scenario)
.WithReportFileName($"testScenario-Report-{DateTime.UtcNow.ToString("yyyy-dd-M--HH-mm-ss")}")
.WithReportFolder("test_reports")
.WithReportFormats(ReportFormat.Html)
.Run();
}
My laptop configuration:
Core i5 10th Gen with 16 Gb RAM, running Windows 10.
At the time of running the load test only VS 2022 was running.
Now I assumed that at 100 RPS it would generate a total of 18k requests for 3 minutes of execution time. The report says different - while it did run for a total of 3 minutes there were only 2057 total requests!
What am I missing here?
How do I get to doing a load test with a higher RPS?
Thanks in advance.
Probably there is a issue with NBomber itself https://github.com/PragmaticFlow/NBomber/issues/488
Also check nbomber log – it could contains a lot of
Error: step unhandled exception: One or more errors occurred. (Too many open files in system
or another errors that indicate that OS limits your load test
I'm facing problem with kestrel server's performance. I have following scenario :
TestClient(JMeter) -> DemoAPI-1(Kestrel) -> DemoAPI-2(IIS)
I'm trying to create a sample application that could get the file content as and when requested.
TestClient(100 Threads) requests to DemoAPI-1 which in turn request to DemoAPI-2. DemoAPI-2 reads a fixed XML file(1 MB max) and returns it's content as a response(In production DemoAPI-2 is not going to be exposed to outside world).
When I tested direct access from TestClient -> DemoAPI-2 I got expected result(good) which is following :
Average : 368ms
Minimum : 40ms
Maximum : 1056ms
Throughput : 40.1/sec
But when I tried to access it through DemoAPI-1 I got following result :
Average : 48232ms
Minimum : 21095ms
Maximum : 49377ms
Throughput : 2.0/sec
As you can see there is a huge difference.I'm not getting even the 10% throughput of DemoAPI-2. I was told has kestrel is more efficient and fast compared to traditional IIS. Also because there is no problem in direct access, I think we can eliminate the possible of problem on DemoAPI-2.
※Code of DemoAPI-1 :
string base64Encoded = null;
var request = new HttpRequestMessage(HttpMethod.Get, url);
var response = await this.httpClient.SendAsync(request, HttpCompletionOption.ResponseContentRead).ConfigureAwait(false);
if (response.StatusCode.Equals(HttpStatusCode.OK))
{
var content = await response.Content.ReadAsByteArrayAsync().ConfigureAwait(false);
base64Encoded = Convert.ToBase64String(content);
}
return base64Encoded;
※Code of DemoAPI-2 :
[HttpGet("Demo2")]
public async Task<IActionResult> Demo2Async(int wait)
{
try
{
if (wait > 0)
{
await Task.Delay(wait);
}
var path = Path.Combine(Directory.GetCurrentDirectory(), "test.xml");
var file = System.IO.File.ReadAllText(path);
return Content(file);
}
catch (System.Exception ex)
{
return StatusCode(500, ex.Message);
}
}
Some additional information :
Both APIs are async.
Both APIs are hosted on different EC2 instances(C5.xlarge Windows Server 2016).
DemoAPI-1(kestrel) is a self-contained API(without reverse proxy)
TestClient(jMeter) is set to 100 thread for this testing.
No other configuration is done for kestrel server as of now.
There are no action filter, middleware or logging that could effect the performance as of now.
Communication is done using SSL on 5001 port.
Wait parameter for DemoAPI2 is set to 0 as of now.
The CPU usage of DEMOAPI-1 is not over 40%.
The problem was due to HttpClient's port exhaustion issue.
I was able to solve this problem by using IHttpClientFactory.
Following article might help someone who faces similar problem.
https://www.stevejgordon.co.uk/httpclient-creation-and-disposal-internals-should-i-dispose-of-httpclient
DEMOAPI-1 performs a non-asynchronous read of the streams:
var bytes = stream.Read(read, 0, DataChunkSize);
while (bytes > 0)
{
buffer += System.Text.Encoding.UTF8.GetString(read, 0, bytes);
// Replace with ReadAsync
bytes = stream.Read(read, 0, DataChunkSize);
}
That can be an issue with throughput on a lot of requests.
Also, I'm not fully aware of why are you not testing the same code with IIS and Kestrel, I would assume you need to make only environmental changes and not the code.
I've encountered a case where the redis pubsub RedisMessageListenerContainer in my spring boot application died with
ERROR .RedisMessageListenerContainer: SubscriptionTask aborted with exception:
org.springframework.dao.QueryTimeoutException: Redis command timed out; nested exception is com.lambdaworks.redis.RedisCommandTimeoutException: Command timed out
at org.springframework.data.redis.connection.lettuce.LettuceExceptionConverter.convert(LettuceExceptionConverter.java:66)
at org.springframework.data.redis.connection.lettuce.LettuceExceptionConverter.convert(LettuceExceptionConverter.java:41)
at org.springframework.data.redis.PassThroughExceptionTranslationStrategy.translate(PassThroughExceptionTranslationStrategy.java:37)
at org.springframework.data.redis.FallbackExceptionTranslationStrategy.translate(FallbackExceptionTranslationStrategy.java:37)
at org.springframework.data.redis.connection.lettuce.LettuceConnection.convertLettuceAccessException(LettuceConnection.java:330)
at org.springframework.data.redis.connection.lettuce.LettuceConnection.subscribe(LettuceConnection.java:3179)
at org.springframework.data.redis.listener.RedisMessageListenerContainer$SubscriptionTask.eventuallyPerformSubscription(RedisMessageListenerContainer.java:790)
at org.springframework.data.redis.listener.RedisMessageListenerContainer$SubscriptionTask.run(RedisMessageListenerContainer.java:746)
at java.lang.Thread.run(Thread.java:748)
Caused by: com.lambdaworks.redis.RedisCommandTimeoutException: Command timed out
at com.lambdaworks.redis.LettuceFutures.await(LettuceFutures.java:113)
at com.lambdaworks.redis.LettuceFutures.awaitOrCancel(LettuceFutures.java:92)
at com.lambdaworks.redis.FutureSyncInvocationHandler.handleInvocation(FutureSyncInvocationHandler.java:63)
at com.lambdaworks.redis.internal.AbstractInvocationHandler.invoke(AbstractInvocationHandler.java:80)
at com.sun.proxy.$Proxy156.subscribe(Unknown Source)
at org.springframework.data.redis.connection.lettuce.LettuceSubscription.doSubscribe(LettuceSubscription.java:63)
at org.springframework.data.redis.connection.util.AbstractSubscription.subscribe(AbstractSubscription.java:142)
at org.springframework.data.redis.connection.lettuce.LettuceConnection.subscribe(LettuceConnection.java:3176)
... 3 common frames omitted
..
I think that shouldn't be an unrecoverable error in the first place because it's a temporary connection issue (and a TransientDataAccessException) but the application apparently needs to deal with exceptions in those case.
Currently this leaves the application in a state that is not acceptable. It merely logs the error but I would either need to kill the application so it gets replaced or better try to restart that container and ideally report via /health that the application is impacted as long as it's not all good.
Is there anything I'm overlooking that is less awkward than either trying to start() the container every x seconds or subclass it and overwrite handleSubscriptionException() and try to act from there? The latter needs much deeper integration with internals than I'd like to have in my code but it's what I so far went with:
RedisMessageListenerContainer container = new RedisMessageListenerContainer() {
#Override
protected void handleSubscriptionException(Throwable ex) {
super.handleSubscriptionException(ex); // don't know what actually happened in here and no way to find out :/
if (ex instanceof RedisConnectionFailureException) {
// handled by super hopefully, don't care
} else if (ex instanceof InterruptedException){
// can ignore those I guess
} else if (ex instanceof TransientDataAccessException || ex instanceof RecoverableDataAccessException) {
// try to restart in those cases?
if (isRunning()) {
logger.error("Connection failure occurred. Restarting subscription task manually due to " + ex, ex);
sleepBeforeRecoveryAttempt();
start(); // best we can do
}
} else {
// otherwise shutdown and hope for the best next time
if (isRunning()) {
logger.warn("Shutting down application due to unknown exception " + ex, ex);
context.close();
}
}
}
};
I try to receive user's PushNotification Channel URI (Windows 10 platform) and for some users application generate exception This operation returned because the timeout.
To handle errors in a task chain I should add a task-based continuation at the end of the chain and handle all errors there (as explained here https://learn.microsoft.com/en-us/windows/uwp/threading-async/asynchronous-programming-in-cpp-universal-windows-platform-apps#handling-errors-in-a-task-chain).
So I did that. But after I call t.get(); system generate Platform::COMException^ exception by it not catch in try-catch block. Why?
There is the code:
{
create_task(PushNotificationChannelManager::CreatePushNotificationChannelForApplicationAsync())
.then([this](PushNotificationChannel^ pnChannel)
{
// ..
// DONE: pnChannel->Uri
// ..
}, task_continuation_context::get_current_winrt_context())
.then([](task<void> t)
{
try
{
t.get(); // <<<< After exec this line app crash!
}
catch (Platform::COMException^ e)
{
OutputDebugString(L"Exception catches!");
}
});
}
There is the full exception message:
Exception thrown at 0x00007FFD9D74A388 in GameName.exe: Microsoft C++
exception: Platform::COMException ^ at memory location 0x000000249A9FEB60.
HRESULT:0x800705B4 This operation returned because the timeout period expired.
WinRT information: This operation returned because the timeout
And Visual Studio throws me to the file c:\Program Files (x86)\Microsoft Visual Studio 14.0\VC\include\exception to block:
[[noreturn]] void _RethrowException() const
{
__ExceptionPtrRethrow(this); // <<<< here everything stoped!
}
UPDATED:
Enviroment:
Visual Studio 2017 (15.7.3)
Project based on cocos2d-x (3.16)
Project Targe Platform Version = 10.0.1493.0
Project Platform Toolset = Visual Studio 2015 (v140)
You can clone cocos2dx project and in MainScene paste code I showed before in onEnter method (or anywhere).
I got a situation when I had crash 100% for me
Uninstall app if it was built before;
Disconnect from the internet;
Build app & launch;
The app will try to detect Channel URI & will crash (but with crash message = WinRT information: The network is not present or not started).
I understand that throwing that exception is normal. But I still cannot understand why when I call t.get() it not catch exception Platform::COMException^
#GetMapping("/")
#ResponseBody
public void getInvoice(#RequestParam String DocumentId,
HttpServletResponse response) {
DocumentDAO documentDAO = null;
try {
documentDAO = service.downloadDocument(DocumentId);
response.setContentType("application/" + documentDAO.getSubtype());
IOUtils.copy(documentDAO.getDocument(), response.getOutputStream());
response.flushBuffer();
documentDAO.getDocument().close();
} catch (IOException e) {
e.printStackTrace();
}
}
The task is to stream pdf document from back-end server (a lot of big documents, up to 200 MB) to the browser via SpringMVC controller. Back-end server outputs document in InputStream - I am copying it to response OutputStream.
IOUtils.copy(documentDAO.getDocument(), response.getOutputStream());
And it works. I just do not like java memory consumption on machine, where this SpringMVC is running.
If it is streaming - why memory consumption increases very high, when customer performs request to this mvc controller?
When big document (i.e. 100 mb) is requested - java heap size increases accordingly.
What I expect is - my java machine should use only some buffer sized amount of memory should not load document to memory, just stream it by.
Is my expectation wrong? Or it is correct and I should do something somehow different?
Thank you in advance.
Here is graph of memory increase when requesting 63MB document
https://i.imgur.com/2fjiHVB.png
and then repeating the request after a while
https://i.imgur.com/Kc77nGM.png
and then GC does its job at the end
https://i.imgur.com/WeIvgoT.png