----- Original Message ----- | From: "Konstantin Kolinko" <knst.koli...@gmail.com> | To: "Tomcat Developers List" <dev@tomcat.apache.org> | Sent: Thursday, November 17, 2011 1:00:45 AM | Subject: Re: Change in behavior for Connector maxKeepAliveRequests in 7.0.21+ | | 2011/11/17 Ryan Morgan <rmor...@vmware.com>: | > | > Hey everyone, | > | > As part of the HTTP connector refactoring that occurred in 7.0.21 | > (see | > http://svn.apache.org/viewvc?diff_format=h&view=revision&revision=1158367) | > the Connector behavior has changed when maxKeepAliveRequests is | > set to 0. In Tomcat 7.0.20 and earlier, a setting of 0 would | > disable keepalives for the connector. In 7.0.21 an up, it's | > enabled and from what I can tell, set to infinite. | > | > Was this change intended? From looking at the documentation, the | > updated code is correct as a value of 0 is not mentioned. | > However, this could cause issues for servers that have | > maxKeepAliveRequests mis-configured to 0 and upgrade to 7.0.21+ | | Just a quick look at Tomcat 6.0 documentation (config/http.html) | reveals that it does not document the value of "0" either. That | specific sentence in Tomcat 6.0 docs was not updated since r420006 | when TC6 development started (5 years ago). So I do not see any | problem here. | | The documented values are -1 for infinity, 1 for disable, 100 by | default. | | The intent of commits around r1158367 was to align implementations | between connectors. It might be that some of them (Bio vs Nio vs Apr) | treated "0" differently. |
Hi Konstantin, Thanks for the reply - I only noticed this change in behavior for the Bio connector. I just wanted to make sure this wasn't something that was overlooked as part of that change. Cheers, -Ryan --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
