Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[jetty-dev] ReadPendingException in Jetty 9.4.8

Hi Everyone,

While running a production setup of Solr with Jetty 9.4.8 we ran into a scenario where the replica node hit's a ReadPendingException . The leader node keeps waiting for the request to complete but eventually times out with a socket timeout exception 

Here's the stack trace excerpt :

java.nio.channels.ReadPendingException: null
at ~[jetty-io-9.4.8.v20171121.jar:9.4.8.v20171121]
at ~[jetty-io-9.4.8.v20171121.jar:9.4.8.v20171121]
at$ ~[jetty-io-9.4.8.v20171121.jar:9.4.8.v20171121]
at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.doProduce( ~[jetty-util-9.4.8.v20171121.jar:9.4.8.v20171121]
at org.eclipse.jetty.util.thread.strategy.EatWhatYouKill.produce( ~[jetty-util-9.4.8.v20171121.jar:9.4.8.v20171121]
at ~[jetty-util-9.4.8.v20171121.jar:9.4.8.v20171121]

I found one existing bug but this seems to have been addressed quite a while ago.

My current analysis was that we're seeing this because of a possible bug in EatWhatYouKill . Is this a known issue or something that needs more debug logs to investigate further?

The current strategy I am planning on taking is upgrading Jetty to 9.4.9+ and see if it fixes the problem. This class has gone through significant changes in Jetty 9.4.9 ( . 


Back to the top