[
https://issues.apache.org/jira/browse/HTTPCLIENT-1478?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14053652#comment-14053652
]
Thijs ten Hoeve edited comment on HTTPCLIENT-1478 at 7/7/14 1:39 PM:
---------------------------------------------------------------------
I too think this bug hasn't been completely fixed.
I have the following trace of an application that blocked indefinitely on a get
request.
{code}
java.lang.Thread.State: RUNNABLE
at java.net.SocketInputStream.socketRead0(Native Method)
at java.net.SocketInputStream.read(SocketInputStream.java:150)
at java.net.SocketInputStream.read(SocketInputStream.java:121)
at sun.security.ssl.InputRecord.readFully(InputRecord.java:465)
at sun.security.ssl.InputRecord.read(InputRecord.java:503)
at sun.security.ssl.SSLSocketImpl.readRecord(SSLSocketImpl.java:954)
- locked <0x00000000e0a6a860> (a java.lang.Object)
at
sun.security.ssl.SSLSocketImpl.performInitialHandshake(SSLSocketImpl.java:1343)
- locked <0x00000000e0a6a870> (a java.lang.Object)
at
sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:1371)
at
sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:1355)
at
org.apache.http.conn.ssl.SSLConnectionSocketFactory.createLayeredSocket(SSLConnectionSocketFactory.java:275)
at
org.apache.http.impl.conn.HttpClientConnectionOperator.upgrade(HttpClientConnectionOperator.java:169)
at
org.apache.http.impl.conn.PoolingHttpClientConnectionManager.upgrade(PoolingHttpClientConnectionManager.java:333)
at
org.apache.http.impl.execchain.MainClientExec.establishRoute(MainClientExec.java:398)
at
org.apache.http.impl.execchain.MainClientExec.execute(MainClientExec.java:219)
at
org.apache.http.impl.execchain.ProtocolExec.execute(ProtocolExec.java:195)
at org.apache.http.impl.execchain.RetryExec.execute(RetryExec.java:86)
at
org.apache.http.impl.execchain.RedirectExec.execute(RedirectExec.java:108)
at
org.apache.http.impl.client.InternalHttpClient.doExecute(InternalHttpClient.java:184)
at
org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:82)
at
org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:106)
{code}
Unfortunately this scenario is very difficult te reproduce.
However I have been able to get similar traces (that didn't block forever since
the server was responsive) in a debugger, and I think I have identified a
possible cause for this.
A connection timeout is not necessarily set on the socket used in the handshake.
This socket is created here:
http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#234
The connection is then opened (With a connectionTimeout) on
http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#239.
However there is no SoTimeout set on the socket (getSoTimeout() returns 0), so
the startHandshake() call in createLayeredSocket
(http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#275)
might block forever.
Note that the Request passed to the execute() call in CloseableHttpClient did
have a SocketTimeout set. ((request.getConfig().getSocketTimeout() == 60000)
was (Author: tth):
I too think this bug hasn't been completely fixed.
I have the following trace of an application that blocked indefinitely on a get
request.
java.lang.Thread.State: RUNNABLE
at java.net.SocketInputStream.socketRead0(Native Method)
at java.net.SocketInputStream.read(SocketInputStream.java:150)
at java.net.SocketInputStream.read(SocketInputStream.java:121)
at sun.security.ssl.InputRecord.readFully(InputRecord.java:465)
at sun.security.ssl.InputRecord.read(InputRecord.java:503)
at sun.security.ssl.SSLSocketImpl.readRecord(SSLSocketImpl.java:954)
- locked <0x00000000e0a6a860> (a java.lang.Object)
at
sun.security.ssl.SSLSocketImpl.performInitialHandshake(SSLSocketImpl.java:1343)
- locked <0x00000000e0a6a870> (a java.lang.Object)
at
sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:1371)
at
sun.security.ssl.SSLSocketImpl.startHandshake(SSLSocketImpl.java:1355)
at
org.apache.http.conn.ssl.SSLConnectionSocketFactory.createLayeredSocket(SSLConnectionSocketFactory.java:275)
at
org.apache.http.impl.conn.HttpClientConnectionOperator.upgrade(HttpClientConnectionOperator.java:169)
at
org.apache.http.impl.conn.PoolingHttpClientConnectionManager.upgrade(PoolingHttpClientConnectionManager.java:333)
at
org.apache.http.impl.execchain.MainClientExec.establishRoute(MainClientExec.java:398)
at
org.apache.http.impl.execchain.MainClientExec.execute(MainClientExec.java:219)
at
org.apache.http.impl.execchain.ProtocolExec.execute(ProtocolExec.java:195)
at org.apache.http.impl.execchain.RetryExec.execute(RetryExec.java:86)
at
org.apache.http.impl.execchain.RedirectExec.execute(RedirectExec.java:108)
at
org.apache.http.impl.client.InternalHttpClient.doExecute(InternalHttpClient.java:184)
at
org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:82)
at
org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:106)
Unfortunately this scenario is very difficult te reproduce.
However I have been able to get similar traces (that didn't block forever since
the server was responsive) in a debugger, and I think I have identified a
possible cause for this.
A connection timeout is not necessarily set on the socket used in the handshake.
This socket is created here:
http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#234
The connection is then opened (With a connectionTimeout) on
http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#239.
However there is no SoTimeout set on the socket (getSoTimeout() returns 0), so
the startHandshake() call in createLayeredSocket
(http://hc.apache.org/httpcomponents-client-4.3.x/httpclient/xref/org/apache/http/conn/ssl/SSLConnectionSocketFactory.html#275)
might block forever.
Note that the Request passed to the execute() call in CloseableHttpClient did
have a SocketTimeout set. ((request.getConfig().getSocketTimeout() == 60000)
> https calls ignore http.socket.timeout during SSL Handshake
> -----------------------------------------------------------
>
> Key: HTTPCLIENT-1478
> URL: https://issues.apache.org/jira/browse/HTTPCLIENT-1478
> Project: HttpComponents HttpClient
> Issue Type: Bug
> Components: HttpConn
> Affects Versions: 4.3 Final, 4.3.1, 4.3.2, 4.3.3
> Environment: All
> Reporter: Jonah Schwartz
> Priority: Minor
> Fix For: 4.3.4
>
>
> https calls ignore http.socket.timeout during SSL Handshake. This can result
> in a https call hanging forever waiting for socket read.
> In both SSLSocketFactory and SSLConnectionSocketFactory,
> sslsock.startHandshake(); is called before socket timeout is set on the
> socket. This means timeout is not respected during the SSL handshake, and the
> thread can hang with a stacktrace that looks like this:
> org.apache.http.impl.client.AbstractHttpClient.doExecute
> org.apache.http.impl.client.DefaultRequestDirector.execute
> org.apache.http.impl.client.DefaultRequestDirector.tryConnect
> org.apache.http.impl.conn.ManagedClientConnectionImpl.open
> org.apache.http.impl.conn.DefaultClientConnectionOperator.openConnection
> org.apache.http.conn.ssl.SSLSocketFactory.connectSocket
> org.apache.http.conn.ssl.SSLSocketFactory.connectSocket
> sun.security.ssl.SSLSocketImpl.startHandshake
> sun.security.ssl.SSLSocketImpl.startHandshake
> sun.security.ssl.SSLSocketImpl.performInitialHandshake
> sun.security.ssl.SSLSocketImpl.readRecord
> sun.security.ssl.InputRecord.read
> sun.security.ssl.InputRecord.readV3Record
> sun.security.ssl.InputRecord.readFully
> java.net.SocketInputStream.read
> java.net.SocketInputStream.socketRead0
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
