RE: FW: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38)
Hi all,
Thanks for the update Christopher.
I tried modifying our application as below i.e by adding a read Listener and
processing the requests after all the data is read (onDataAvailable()) .
But I can still see the StackOverFlow Error printed .
AsyncContext asyncContext = req.startAsync( req, resp );
ServletInputStream input = req.getInputStream();
ReadListenerImpl listener = new ReadListenerImpl( input,
asyncContext, req, resp, map, myExecutor, this );
input.setReadListener( listener );
. . .
public class ReadListenerImpl implements ReadListener
{
.
.
.
@Override
public void onDataAvailable() throws IOException
{
int len = -1;
byte b[] = new byte[4 * 1024];
while( input.isReady() &&
!input.isFinished() )
{
len = input.read( b );
if( len > 0 )
{
String data = new
String( b, 0, len );
sb.append( data );
}
}
}
@Override
public void onError(Throwable throwable)
{
asyncContext.complete();
throwable.printStackTrace();
}
@Override
public void onAllDataRead() throws IOException
{
String jsonData = sb.toString();
.
.
.
}
}
Unfortunately I am still not able to reproduce this exception with a test code
, I am trying .
Thanks and Regards
Arshiya Shariff
-Original Message-
From: Christopher Schultz
Sent: Wednesday, October 21, 2020 7:53 PM
To: users@tomcat.apache.org
Subject: Re: FW: HTTP2: memory filled up fast on increasing the connections to
1000/2000 (Embedded tomcat 9.0.38)
Arshiya,
On 10/21/20 00:34, Arshiya Shariff wrote:
> Hi,
>
> Christopher, Please find the answer in-line:
> How... exactly?
> private String getRequestBody(HttpServletRequest request) throws IOException
> {
> StringBuilder sb = new StringBuilder();
> BufferedReader reader = request.getReader();
> try
> {
> String line;
> while( ( line = reader.readLine() ) != null )
> {
> sb.append( line ).append( '\n' );
Note that this may modify the incoming request. Are you sure you want to return
a value which does not match the exact POST body?
> }
> }
> finally
> {
> reader.close();
> }
> return sb.toString();
> }
>
Is that method run from within the asynchronous context or before you begin
async processing? I'm not an expert at servlet-async, but I think you should be
reading the request entirely before entering the async context. Reading the
request from async may cause problems.
Instead of using blocking reads of the request bbody in asynchronous mode, you
should do this:
request.getInputStream().setReadListener(new ReadListener() {
public void onoAllDataRead() { ... }
public void onDataAvailable() { ... }
public void onError(Throwable t) { ... } });
> I am trying to reproduce the StackOverflowError with a sample
> application , Once it is reproduced I will share it across.
See https://www.slideshare.net/SimoneBordet/servlet-31-async-io
Specifically slides 43-53.
-chris
> -Original Message-
> From: Christopher Schultz
> Sent: Thursday, October 15, 2020 12:01 AM
> To: users@tomcat.apache.org
> Subject: Re: FW: HTTP2: memory filled up fast on increasing the
> connections to 1000/2000 (Embedded tomcat 9.0.38)
>
> Arshiya,
>
> On 10/14/20 01:23, Arshiya Shariff wrote:
>> Please find the answers in-line Mark.
>>
>> Http2 requests with message payload of 34KB are pumped from JMeter
>> at
>> 20 TPS with 700 connections to an application with
Re: FW: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38)
Arshiya,
On 10/21/20 00:34, Arshiya Shariff wrote:
> Hi,
>
> Christopher, Please find the answer in-line:
> How... exactly?
> private String getRequestBody(HttpServletRequest request) throws IOException
> {
> StringBuilder sb = new StringBuilder();
> BufferedReader reader = request.getReader();
> try
> {
> String line;
> while( ( line = reader.readLine() ) != null )
> {
> sb.append( line ).append( '\n' );
Note that this may modify the incoming request. Are you sure you want to
return a value which does not match the exact POST body?
> }
> }
> finally
> {
> reader.close();
> }
> return sb.toString();
> }
>
Is that method run from within the asynchronous context or before you
begin async processing? I'm not an expert at servlet-async, but I think
you should be reading the request entirely before entering the async
context. Reading the request from async may cause problems.
Instead of using blocking reads of the request bbody in asynchronous
mode, you should do this:
request.getInputStream().setReadListener(new ReadListener() {
public void onoAllDataRead() { ... }
public void onDataAvailable() { ... }
public void onError(Throwable t) { ... }
});
> I am trying to reproduce the StackOverflowError with a sample
> application , Once it is reproduced I will share it across.
See https://www.slideshare.net/SimoneBordet/servlet-31-async-io
Specifically slides 43-53.
-chris
> -Original Message-
> From: Christopher Schultz
> Sent: Thursday, October 15, 2020 12:01 AM
> To: users@tomcat.apache.org
> Subject: Re: FW: HTTP2: memory filled up fast on increasing the connections
> to 1000/2000 (Embedded tomcat 9.0.38)
>
> Arshiya,
>
> On 10/14/20 01:23, Arshiya Shariff wrote:
>> Please find the answers in-line Mark.
>>
>> Http2 requests with message payload of 34KB are pumped from JMeter at
>> 20 TPS with 700 connections to an application with Embedded tomcat
>> - 9.0.39 (max-Threads : 200, all other values are the tomcat
>> defaults)
>>
>>> What does that URL do with the POSTed content? Ignore it? Read it
>>> from an InputStream? Read it via getParameter()?
>>
>> The posted content is read via BufferedReader reader =
>> request.getReader() and processed asynchronously
> How... exactly?
>> Is JMeter run on the same machine as Tomcat?
>> JMeter is run from a different machine.
>>
>> Do you use the JMeter GUI or the command line?
>> Launched via Command line (JMeter heap increased to 10 GB )
>>
>> What are the specs of the server(s) being used?
>> The server is a VM with 12 CPUs and 120 GB RAM
>>
>> Please let us know if you require more details.
>
> This would probabyl be easier if you'd just provide a test-case: a sample
> (simple!) web application which reproduces what you are reporting.
>
> -chris
>
>> -Original Message-
>> From: Mark Thomas
>> Sent: Monday, October 12, 2020 7:28 PM
>> To: users@tomcat.apache.org
>> Subject: Re: HTTP2: memory filled up fast on increasing the
>> connections to 1000/2000 (Embedded tomcat 9.0.38)
>>
>> On 12/10/2020 08:02, Arshiya Shariff wrote:
>>> Hi Mark ,
>>>
>>> The issue is reproduced with version 9.0.39 as well. Max threads in Tomcat
>>> is 200.
>>>
>>> Please find the case:
>>> Client:JMeter 5.2.1 (With http2 plugin)
>>> TPS: around 20
>>> No of users from JMeter : 700
>>> Message payload size: 6 KB to 34 KB
>>> Loop: Infinite
>>> We let the loop run infinitely and see the java.lang.StackOverflowError
>>> trace printed multiple times in the log within few minutes of starting the
>>> test.
>>
>> POSTing to what URL?
>>
>> What does that URL do with the POSTed content? Ignore it? Read it from an
>> InputStream? Read it via getParameter()?
>>
>> Is JMeter run on the same machine as Tomcat?
>>
>> Do you use the JMeter GUI or the command line?
>>
>> What are the specs of the server(s) being used?
>>
>> You need to provide the exact steps to recreate this issue on a clean
>> install of Tomcat 9.0.39 as provided by the ASF.
>>
>> Mark
>>
>>
>>> Please help us with this . What is the impact of StackOverflowError ?
>>>
>>> Thanks and Regards
>>> Arshiya Shariff
>>>
>>> -Original Message-
>>> From: Mark Thomas
>>> Sent: Friday, October 9, 2020 5:31 PM
>>> To: users@tomcat.apache.org
>>> Subject: Re: HTTP2: memory filled up fast on increasing the
>>> connections to 1000/2000 (Embedded tomcat 9.0.38)
>>>
>>> On 09/10/2020 12:32, Arshiya Shariff wrote:
Hi,
Mark , with the test runs that I performed over clean 9.0.x branch I was
not able to reproduce this.
>>>
>>> Good. But I'd really like to understand why...
>>>
But with 9.0.38 and the jars built from 9.0.x with hash:
c8ec2d4cde3a31b0e9df9a30e7915d77ba725545 ,
RE: FW: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38)
Hi,
Christopher, Please find the answer in-line:
How... exactly?
private String getRequestBody(HttpServletRequest request) throws IOException
{
StringBuilder sb = new StringBuilder();
BufferedReader reader = request.getReader();
try
{
String line;
while( ( line = reader.readLine() ) != null )
{
sb.append( line ).append( '\n' );
}
}
finally
{
reader.close();
}
return sb.toString();
}
I am trying to reproduce the StackOverflowError with a sample application ,
Once it is reproduced I will share it across.
Thanks and Regards
Arshiya Shariff
-Original Message-
From: Christopher Schultz
Sent: Thursday, October 15, 2020 12:01 AM
To: users@tomcat.apache.org
Subject: Re: FW: HTTP2: memory filled up fast on increasing the connections to
1000/2000 (Embedded tomcat 9.0.38)
Arshiya,
On 10/14/20 01:23, Arshiya Shariff wrote:
> Please find the answers in-line Mark.
>
> Http2 requests with message payload of 34KB are pumped from JMeter at
> 20 TPS with 700 connections to an application with Embedded tomcat
> - 9.0.39 (max-Threads : 200, all other values are the tomcat
> defaults)
>
>> What does that URL do with the POSTed content? Ignore it? Read it
>> from an InputStream? Read it via getParameter()?
>
> The posted content is read via BufferedReader reader =
> request.getReader() and processed asynchronously
How... exactly?
> Is JMeter run on the same machine as Tomcat?
> JMeter is run from a different machine.
>
> Do you use the JMeter GUI or the command line?
> Launched via Command line (JMeter heap increased to 10 GB )
>
> What are the specs of the server(s) being used?
> The server is a VM with 12 CPUs and 120 GB RAM
>
> Please let us know if you require more details.
This would probabyl be easier if you'd just provide a test-case: a sample
(simple!) web application which reproduces what you are reporting.
-chris
> -Original Message-
> From: Mark Thomas
> Sent: Monday, October 12, 2020 7:28 PM
> To: users@tomcat.apache.org
> Subject: Re: HTTP2: memory filled up fast on increasing the
> connections to 1000/2000 (Embedded tomcat 9.0.38)
>
> On 12/10/2020 08:02, Arshiya Shariff wrote:
>> Hi Mark ,
>>
>> The issue is reproduced with version 9.0.39 as well. Max threads in Tomcat
>> is 200.
>>
>> Please find the case:
>> Client:JMeter 5.2.1 (With http2 plugin)
>> TPS: around 20
>> No of users from JMeter : 700
>> Message payload size: 6 KB to 34 KB
>> Loop: Infinite
>> We let the loop run infinitely and see the java.lang.StackOverflowError
>> trace printed multiple times in the log within few minutes of starting the
>> test.
>
> POSTing to what URL?
>
> What does that URL do with the POSTed content? Ignore it? Read it from an
> InputStream? Read it via getParameter()?
>
> Is JMeter run on the same machine as Tomcat?
>
> Do you use the JMeter GUI or the command line?
>
> What are the specs of the server(s) being used?
>
> You need to provide the exact steps to recreate this issue on a clean install
> of Tomcat 9.0.39 as provided by the ASF.
>
> Mark
>
>
>> Please help us with this . What is the impact of StackOverflowError ?
>>
>> Thanks and Regards
>> Arshiya Shariff
>>
>> -Original Message-
>> From: Mark Thomas
>> Sent: Friday, October 9, 2020 5:31 PM
>> To: users@tomcat.apache.org
>> Subject: Re: HTTP2: memory filled up fast on increasing the
>> connections to 1000/2000 (Embedded tomcat 9.0.38)
>>
>> On 09/10/2020 12:32, Arshiya Shariff wrote:
>>> Hi,
>>>
>>> Mark , with the test runs that I performed over clean 9.0.x branch I was
>>> not able to reproduce this.
>>
>> Good. But I'd really like to understand why...
>>
>>> But with 9.0.38 and the jars built from 9.0.x with hash:
>>> c8ec2d4cde3a31b0e9df9a30e7915d77ba725545 , with 700 or 1000 users
>>> (connections) and on sending 1000 Requests per second (or even lesser) ,
>>> payload of 16K from JMeter I can see that this Exception occurs within few
>>> minutes of starting the test . The maxThreads configured in tomcat is 200 .
>>>
>>> How often do you see these errors in your test run?
>>> Randomly, at times 2 or 3 such traces.
>>
>> OK. Definitely a timing issue then.
>>
>>> Do you have the other end of that stack trace?
>>> It is only the two lines that is recursively printed till the end about
>>> ~500 times in one trace :
>>> at
>>> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1511)
>>> at
>>> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHan
>>> d
>>> l
>>> er.completed(SocketWrapperBase.java:1100)
>>
>> Doesn't tell me much unfortunately.
>>
>>> I see the trace starting with :
>>>
Re: FW: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38)
Arshiya, On 10/14/20 01:23, Arshiya Shariff wrote: > Please find the answers in-line Mark. > > Http2 requests with message payload of 34KB are pumped from JMeter > at 20 TPS with 700 connections to an application with Embedded tomcat > - 9.0.39 (max-Threads : 200, all other values are the tomcat > defaults) > >> What does that URL do with the POSTed content? Ignore it? Read it >> from an InputStream? Read it via getParameter()? > > The posted content is read via BufferedReader reader = > request.getReader() and processed asynchronously How... exactly? > Is JMeter run on the same machine as Tomcat? > JMeter is run from a different machine. > > Do you use the JMeter GUI or the command line? > Launched via Command line (JMeter heap increased to 10 GB ) > > What are the specs of the server(s) being used? > The server is a VM with 12 CPUs and 120 GB RAM > > Please let us know if you require more details. This would probabyl be easier if you'd just provide a test-case: a sample (simple!) web application which reproduces what you are reporting. -chris > -Original Message- > From: Mark Thomas > Sent: Monday, October 12, 2020 7:28 PM > To: users@tomcat.apache.org > Subject: Re: HTTP2: memory filled up fast on increasing the connections to > 1000/2000 (Embedded tomcat 9.0.38) > > On 12/10/2020 08:02, Arshiya Shariff wrote: >> Hi Mark , >> >> The issue is reproduced with version 9.0.39 as well. Max threads in Tomcat >> is 200. >> >> Please find the case: >> Client:JMeter 5.2.1 (With http2 plugin) >> TPS: around 20 >> No of users from JMeter : 700 >> Message payload size: 6 KB to 34 KB >> Loop: Infinite >> We let the loop run infinitely and see the java.lang.StackOverflowError >> trace printed multiple times in the log within few minutes of starting the >> test. > > POSTing to what URL? > > What does that URL do with the POSTed content? Ignore it? Read it from an > InputStream? Read it via getParameter()? > > Is JMeter run on the same machine as Tomcat? > > Do you use the JMeter GUI or the command line? > > What are the specs of the server(s) being used? > > You need to provide the exact steps to recreate this issue on a clean install > of Tomcat 9.0.39 as provided by the ASF. > > Mark > > >> Please help us with this . What is the impact of StackOverflowError ? >> >> Thanks and Regards >> Arshiya Shariff >> >> -Original Message- >> From: Mark Thomas >> Sent: Friday, October 9, 2020 5:31 PM >> To: users@tomcat.apache.org >> Subject: Re: HTTP2: memory filled up fast on increasing the >> connections to 1000/2000 (Embedded tomcat 9.0.38) >> >> On 09/10/2020 12:32, Arshiya Shariff wrote: >>> Hi, >>> >>> Mark , with the test runs that I performed over clean 9.0.x branch I was >>> not able to reproduce this. >> >> Good. But I'd really like to understand why... >> >>> But with 9.0.38 and the jars built from 9.0.x with hash: >>> c8ec2d4cde3a31b0e9df9a30e7915d77ba725545 , with 700 or 1000 users >>> (connections) and on sending 1000 Requests per second (or even lesser) , >>> payload of 16K from JMeter I can see that this Exception occurs within few >>> minutes of starting the test . The maxThreads configured in tomcat is 200 . >>> >>> How often do you see these errors in your test run? >>> Randomly, at times 2 or 3 such traces. >> >> OK. Definitely a timing issue then. >> >>> Do you have the other end of that stack trace? >>> It is only the two lines that is recursively printed till the end about >>> ~500 times in one trace : >>> at >>> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1511) >>> at >>> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHand >>> l >>> er.completed(SocketWrapperBase.java:1100) >> >> Doesn't tell me much unfortunately. >> >>> I see the trace starting with : >>> Exception in thread "http-nio-x.y.z-1090-exec-107" >>> java.lang.StackOverflowError >>> at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:446) >>> at org.apache.tomcat.util.net.NioChannel.read(NioChannel.java:174) >>> at >>> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1468) >>> at >>> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHand >>> l >>> er.completed(SocketWrapperBase.java:1100) >>> >>> (OR) >>> >>> Exception in thread "http-nio-x.y.z-1090-exec-87" >>> java.lang.StackOverflowError >>> at sun.nio.ch.IOVecWrapper.get(IOVecWrapper.java:96) >>> at sun.nio.ch.IOUtil.read(IOUtil.java:240) >>> at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:440) >>> at org.apache.tomcat.util.net.NioChannel.read(NioChannel.java:174) >>> at >>> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1468) >>> at >>>
FW: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38)
Hi , Please find the answers in-line Mark. Http2 requests with message payload of 34KB are pumped from JMeter at 20 TPS with 700 connections to an application with Embedded tomcat - 9.0.39 (max-Threads : 200, all other values are the tomcat defaults) What does that URL do with the POSTed content? Ignore it? Read it from an InputStream? Read it via getParameter()? The posted content is read via BufferedReader reader = request.getReader() and processed asynchronously. Is JMeter run on the same machine as Tomcat? JMeter is run from a different machine. Do you use the JMeter GUI or the command line? Launched via Command line (JMeter heap increased to 10 GB ) What are the specs of the server(s) being used? The server is a VM with 12 CPUs and 120 GB RAM Please let us know if you require more details. Thanks and Regards Arshiya Shariff -Original Message- From: Mark Thomas Sent: Monday, October 12, 2020 7:28 PM To: users@tomcat.apache.org Subject: Re: HTTP2: memory filled up fast on increasing the connections to 1000/2000 (Embedded tomcat 9.0.38) On 12/10/2020 08:02, Arshiya Shariff wrote: > Hi Mark , > > The issue is reproduced with version 9.0.39 as well. Max threads in Tomcat is > 200. > > Please find the case: > Client:JMeter 5.2.1 (With http2 plugin) > TPS: around 20 > No of users from JMeter : 700 > Message payload size: 6 KB to 34 KB > Loop: Infinite > We let the loop run infinitely and see the java.lang.StackOverflowError trace > printed multiple times in the log within few minutes of starting the test. POSTing to what URL? What does that URL do with the POSTed content? Ignore it? Read it from an InputStream? Read it via getParameter()? Is JMeter run on the same machine as Tomcat? Do you use the JMeter GUI or the command line? What are the specs of the server(s) being used? You need to provide the exact steps to recreate this issue on a clean install of Tomcat 9.0.39 as provided by the ASF. Mark > Please help us with this . What is the impact of StackOverflowError ? > > Thanks and Regards > Arshiya Shariff > > -Original Message- > From: Mark Thomas > Sent: Friday, October 9, 2020 5:31 PM > To: users@tomcat.apache.org > Subject: Re: HTTP2: memory filled up fast on increasing the > connections to 1000/2000 (Embedded tomcat 9.0.38) > > On 09/10/2020 12:32, Arshiya Shariff wrote: >> Hi, >> >> Mark , with the test runs that I performed over clean 9.0.x branch I was not >> able to reproduce this. > > Good. But I'd really like to understand why... > >> But with 9.0.38 and the jars built from 9.0.x with hash: >> c8ec2d4cde3a31b0e9df9a30e7915d77ba725545 , with 700 or 1000 users >> (connections) and on sending 1000 Requests per second (or even lesser) , >> payload of 16K from JMeter I can see that this Exception occurs within few >> minutes of starting the test . The maxThreads configured in tomcat is 200 . >> >> How often do you see these errors in your test run? >> Randomly, at times 2 or 3 such traces. > > OK. Definitely a timing issue then. > >> Do you have the other end of that stack trace? >> It is only the two lines that is recursively printed till the end about >> ~500 times in one trace : >> at >> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1511) >> at >> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHand >> l >> er.completed(SocketWrapperBase.java:1100) > > Doesn't tell me much unfortunately. > >> I see the trace starting with : >> Exception in thread "http-nio-x.y.z-1090-exec-107" >> java.lang.StackOverflowError >> at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:446) >> at org.apache.tomcat.util.net.NioChannel.read(NioChannel.java:174) >> at >> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1468) >> at >> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHand >> l >> er.completed(SocketWrapperBase.java:1100) >> >> (OR) >> >> Exception in thread "http-nio-x.y.z-1090-exec-87" >> java.lang.StackOverflowError >> at sun.nio.ch.IOVecWrapper.get(IOVecWrapper.java:96) >> at sun.nio.ch.IOUtil.read(IOUtil.java:240) >> at sun.nio.ch.SocketChannelImpl.read(SocketChannelImpl.java:440) >> at org.apache.tomcat.util.net.NioChannel.read(NioChannel.java:174) >> at >> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1468) >> at >> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHandler.completed(SocketWrapperBase.java:1100) >> at >> org.apache.tomcat.util.net.NioEndpoint$NioSocketWrapper$NioOperationState.run(NioEndpoint.java:1511) >> at >> org.apache.tomcat.util.net.SocketWrapperBase$VectoredIOCompletionHandler.completed(SocketWrapperBase.java:1100) >> . >> . >> . >> .
