You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ws.apache.org by "Michael Haeusler (JIRA)" <ji...@apache.org> on 2018/12/11 16:41:00 UTC

[jira] [Created] (WSS-638) out of memory when validating XML-Signature of an encypted SOAP-Message with large attachment

Michael Haeusler created WSS-638:
------------------------------------

             Summary: out of memory when validating XML-Signature of an encypted SOAP-Message with large attachment
                 Key: WSS-638
                 URL: https://issues.apache.org/jira/browse/WSS-638
             Project: WSS4J
          Issue Type: Bug
          Components: WSS4J Axis Integration
    Affects Versions: 2.2.1
            Reporter: Michael Haeusler
            Assignee: Colm O hEigeartaigh


I have to process a SOAP message with a binary attachment, where XML-Encryption and XML-Signature are applied to the SOAP-Header and attachment  (SOAP-with-Attachments)

everything works just fine when the attachment size fits into the JVM memory, however the attachment can get larger than 1GB and this results in the following stacktrace:

 
{code:java}
java.lang.OutOfMemoryError: Java heap space
at java.util.Arrays.copyOf(Arrays.java:3236)
at java.io.ByteArrayOutputStream.grow(ByteArrayOutputStream.java:118)
at java.io.ByteArrayOutputStream.ensureCapacity(ByteArrayOutputStream.java:93)
at java.io.ByteArrayOutputStream.write(ByteArrayOutputStream.java:153)
at com.sun.crypto.provider.GaloisCounterMode.decrypt(GaloisCounterMode.java:498)
at com.sun.crypto.provider.CipherCore.update(CipherCore.java:772)
at com.sun.crypto.provider.CipherCore.update(CipherCore.java:671)
at com.sun.crypto.provider.AESCipher.engineUpdate(AESCipher.java:380)
at javax.crypto.Cipher.update(Cipher.java:1831)
at javax.crypto.CipherInputStream.getMoreData(CipherInputStream.java:132)
at javax.crypto.CipherInputStream.read(CipherInputStream.java:239)
at org.apache.wss4j.common.util.AttachmentUtils$1.read(AttachmentUtils.java:528)
at java.io.BufferedInputStream.fill(BufferedInputStream.java:246)
at java.io.BufferedInputStream.read1(BufferedInputStream.java:286)
at java.io.BufferedInputStream.read(BufferedInputStream.java:345)
at java.io.FilterInputStream.read(FilterInputStream.java:133)
at java.io.FilterInputStream.read(FilterInputStream.java:107)
at org.apache.wss4j.dom.transform.AttachmentContentSignatureTransform.processAttachment(AttachmentContentSignatureTransform.java:211)
at org.apache.wss4j.dom.transform.AttachmentContentSignatureTransform.transform(AttachmentContentSignatureTransform.java:115)
at org.apache.jcp.xml.dsig.internal.dom.DOMTransform.transform(DOMTransform.java:159)
at org.apache.jcp.xml.dsig.internal.dom.DOMReference.transform(DOMReference.java:443)
at org.apache.jcp.xml.dsig.internal.dom.DOMReference.validate(DOMReference.java:371)
at org.apache.jcp.xml.dsig.internal.dom.DOMXMLSignature.validate(DOMXMLSignature.java:265)
at org.apache.wss4j.dom.processor.SignatureProcessor.verifyXMLSignature(SignatureProcessor.java:373)
at org.apache.wss4j.dom.processor.SignatureProcessor.handleToken(SignatureProcessor.java:233)
at org.apache.wss4j.dom.engine.WSSecurityEngine.processSecurityHeader(WSSecurityEngine.java:340)
at org.apache.wss4j.dom.engine.WSSecurityEngine.processSecurityHeader(WSSecurityEngine.java:251)
{code}
I have debugged the library and found that there is a fundamental problem in 
 org.apache.wss4j.dom.transform.AttachmentContentSignatureTransform
{code:java}
protected Data processAttachment(XMLCryptoContext context, OutputStream os, String attachmentUri, Attachment attachment) throws TransformException {
   try {
      //try to reuse the inputStream in the hope that the provided inputStream is backed by a disk storage
      InputStream inputStream = attachment.getSourceStream();
      if (!inputStream.markSupported()) {
          inputStream = new BufferedInputStream(inputStream);
      }
      inputStream.mark(Integer.MAX_VALUE); //we can process at maximum 2G with the standard jdk streams
{code}
there are two problems with this code:
 # the inputStream.mark() causes the BufferedInputStream to grow up to 2GB

I have worked around this by providing a custom BufferedInputStream that uses a disk storage backend, so that no memory is used. this fixes the problem for the opposite direction where the same SOAP messages are created.


 # for validation case there is no way to create a workaround, because my attachment stream is automatically wrapped in an anonymous CipherInputStream in org.apache.wss4j.dom.util.EncryptionUtils.decryptAttachement(String, String, Element,SecretKey, String, CallbackHandler)
which causes a new BufferedInputStream to be wrapped around it, as shown in the source code above.



 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@ws.apache.org
For additional commands, e-mail: dev-help@ws.apache.org