Jetty

Requests hang when increasing the number of BlockingChannelConnector acceptors

Details

  • Type: Bug Bug
  • Status: Open Open
  • Priority: Major Major
  • Resolution: Unresolved
  • Affects Version/s: 7.5.4
  • Fix Version/s: None
  • Component/s: None
  • Labels:
    None
  • Environment:
    Reproduced on OS X 10.7.2 and Ubuntu Server 10.10
  • Number of attachments :
    0

Description

Requests start hanging as the number of acceptors reaches or passes the number of available processors.

Reproducible test case: https://github.com/collinvandyck/jetty-issue

Activity

Ascending order - Click to sort in descending order
Hide
Permalink
Collin VanDyck added a comment -

This test seems to be easily reproducible for runtimes which report >= 4 processors. Tests on virtual machines reporting < 4 processors passed the test.

Show
Collin VanDyck added a comment - This test seems to be easily reproducible for runtimes which report >= 4 processors. Tests on virtual machines reporting < 4 processors passed the test.
Hide
Permalink
Joakim Erdfelt added a comment -

How do you determine the number of processors that a runtime reports?

Is this a representative system of 8 processors?

$ lsb_release -a
LSB Version: core-2.0-amd64:core-2.0-noarch:core-3.0-amd64:core-3.0-noarch:core-3.1-amd64:core-3.1-noarch:core-3.2-amd64:core-3.2-noarch:core-4.0-amd64:core-4.0-noarch
Distributor ID: Ubuntu
Description: Ubuntu 11.10
Release: 11.10
Codename: oneiric

$ java -version
java version "1.6.0_27"
Java(TM) SE Runtime Environment (build 1.6.0_27-b07)
Java HotSpot(TM) 64-Bit Server VM (build 20.2-b06, mixed mode)

$ cat /proc/cpuinfo | grep -E "^processor"
processor : 0
processor : 1
processor : 2
processor : 3
processor : 4
processor : 5
processor : 6
processor : 7

Show
Joakim Erdfelt added a comment - How do you determine the number of processors that a runtime reports? Is this a representative system of 8 processors? $ lsb_release -a LSB Version: core-2.0-amd64:core-2.0-noarch:core-3.0-amd64:core-3.0-noarch:core-3.1-amd64:core-3.1-noarch:core-3.2-amd64:core-3.2-noarch:core-4.0-amd64:core-4.0-noarch Distributor ID: Ubuntu Description: Ubuntu 11.10 Release: 11.10 Codename: oneiric $ java -version java version "1.6.0_27" Java(TM) SE Runtime Environment (build 1.6.0_27-b07) Java HotSpot(TM) 64-Bit Server VM (build 20.2-b06, mixed mode) $ cat /proc/cpuinfo | grep -E "^processor" processor : 0 processor : 1 processor : 2 processor : 3 processor : 4 processor : 5 processor : 6 processor : 7
Hide
Permalink
Collin VanDyck added a comment -

Sorry, I mean the JVM runtime.

Runtime.getRuntime().availableProcessors()
Show
Collin VanDyck added a comment - Sorry, I mean the JVM runtime. Runtime .getRuntime().availableProcessors()
Hide
Permalink
Joakim Erdfelt added a comment -

Test case provided at github link above replicates problem.

Results:

BlockingChannelConnector = FAILURE at numProcessors - 1
SelectChannelConnector = success at numProcessors * 2
SocketConnector = FAILURE at numProcessors + 1
SslSelectChannelConnector = success at numProcessors * 2
SslSocketConnector = FAILURE at numProcessors + 1

Show
Joakim Erdfelt added a comment - Test case provided at github link above replicates problem. Results: BlockingChannelConnector = FAILURE at numProcessors - 1 SelectChannelConnector = success at numProcessors * 2 SocketConnector = FAILURE at numProcessors + 1 SslSelectChannelConnector = success at numProcessors * 2 SslSocketConnector = FAILURE at numProcessors + 1
Hide
Permalink
Joakim Erdfelt added a comment -

Interestingly, if you start with 8 acceptors on BlockingChannelConnector it works for 8, but then fails for 9 (like SocketConnector does)

Show
Joakim Erdfelt added a comment - Interestingly, if you start with 8 acceptors on BlockingChannelConnector it works for 8, but then fails for 9 (like SocketConnector does)

People

  • Assignee:
    Joakim Erdfelt
    Reporter:
    Collin VanDyck
Vote (0)
Watch (2)

Dates

  • Created:
    Updated: