After upgrading to Crowd 3.0.1 from version 2.11.1 loading images, CSS, and JS files stopped working. It seems that the response header indicates that GZIP compression is used, but the content is considered invalid by the browser.
At the same time the following error is logged:
WARN [webresource] error in `Content.writeTo` for com.atlassian.plugins.atlassian-nps-plugin:nps-btf-admin-resources/btf-admin.js
java.lang.RuntimeException: com.atlassian.plugin.servlet.DownloadException: Unable to stream to the output
at com.atlassian.plugin.webresource.impl.helpers.ResourceServingHelpers$6.writeTo(ResourceServingHelpers.java:360)
at com.atlassian.plugin.webresource.impl.helpers.ResourceServingHelpers$2.lambda$writeTo$0(ResourceServingHelpers.java:184)
at com.atlassian.plugin.cache.filecache.impl.PassThroughCache.cache(PassThroughCache.java:18)
at com.atlassian.plugin.webresource.impl.helpers.ResourceServingHelpers$2.writeTo(ResourceServingHelpers.java:184)
at com.atlassian.plugin.webresource.impl.helpers.ResourceServingHelpers$8.writeTo(ResourceServingHelpers.java:384)
[..]
Caused by: com.atlassian.plugin.servlet.DownloadException: Unable to stream to the output
at com.atlassian.plugin.webresource.transformer.TransformerUtils.transformAndStreamResource(TransformerUtils.java:44)
at com.atlassian.plugin.webresource.transformer.CharSequenceDownloadableResource.streamResource(CharSequenceDownloadableResource.java:42)
at com.atlassian.plugin.webresource.impl.helpers.ResourceServingHelpers$6.writeTo(ResourceServingHelpers.java:358)
... 114 more
Caused by: org.apache.catalina.connector.ClientAbortException: java.net.SocketException: Broken pipe (Write failed)
at org.apache.catalina.connector.OutputBuffer.doFlush(OutputBuffer.java:367)
at org.apache.catalina.connector.OutputBuffer.flush(OutputBuffer.java:331)
at org.apache.catalina.connector.CoyoteOutputStream.flush(CoyoteOutputStream.java:101)
at sun.nio.cs.StreamEncoder.implFlush(StreamEncoder.java:297)
at sun.nio.cs.StreamEncoder.flush(StreamEncoder.java:141)
at java.io.OutputStreamWriter.flush(OutputStreamWriter.java:229)
at org.apache.commons.io.IOUtils.copy(IOUtils.java:2128)
at org.apache.commons.io.IOUtils.copy(IOUtils.java:2158)
at com.atlassian.plugin.webresource.transformer.TransformerUtils.transformAndStreamResource(TransformerUtils.java:40)
... 116 more
Caused by: java.net.SocketException: Broken pipe (Write failed)
at java.net.SocketOutputStream.socketWrite0(Native Method)
at java.net.SocketOutputStream.socketWrite(SocketOutputStream.java:111)
at java.net.SocketOutputStream.write(SocketOutputStream.java:155)
at org.apache.coyote.http11.InternalOutputBuffer.realWriteBytes(InternalOutputBuffer.java:215)
at org.apache.tomcat.util.buf.ByteChunk.flushBuffer(ByteChunk.java:480)
at org.apache.coyote.http11.InternalOutputBuffer.flush(InternalOutputBuffer.java:119)
at org.apache.coyote.http11.AbstractHttp11Processor.action(AbstractHttp11Processor.java:800)
at org.apache.coyote.Response.action(Response.java:172)
at org.apache.catalina.connector.OutputBuffer.doFlush(OutputBuffer.java:363)
... 124 more
Note that connecting using localhost does work because GZIP compression is then automatically disabled.
I found a workaround by disabling compression as described here:
https://confluence.atlassian.com/crowdkb/how-to-disable-http-compression-on-crowd-2-9-and-above-856714638.html
Why did this stop working? What is the nps plugin? And why is it messing with output streams?