Re: ocsp.t failing
On Wed, Jan 12, 2022 at 11:09 AM Joe Orton wrote: > > > > I wonder if this is due to change in the OpenSSL version in Ubuntu? > > Appears not :) Big thanks to Yann for fixing it! Actually yes :) An OpenSSL change which caused a modification of the perl IO-Socket-SSL module [1] which caused a different error message returned for rejected handshakes. Most credits go to DimStar on #httpd though, who pinged me when (s)he saw that we disabled the test for "unknwon reasons" and that I was fighting with travis to show me something. (S)he bisected and pointed me to the offending commit in IO-Socket-SSL thus allowing me to upgrade the module locally and reproduce, the fix was then straightforward.. So many thanks to DimStar! [1] https://github.com/noxxi/p5-io-socket-ssl/commit/0c22923961535a7602a287acb8f65a88c7705a86
Re: ocsp.t failing
On Wed, Jan 05, 2022 at 12:11:50PM +, Joe Orton wrote: > On Sat, Dec 25, 2021 at 10:56:59AM +0100, Christophe JAILLET wrote: > > Hi, > > > > the last travis jobs have failed on trunk because of ocsp.t. > > > > Unless I missed something, the only change with previous successful build is > > r1896361: > > > > https://github.com/apache/httpd/commit/442b4b167f19e13df918402a7af28fe4a50c2730 > > > > I can't see any potential link. > > I wonder if this is due to change in the OpenSSL version in Ubuntu? Appears not :) Big thanks to Yann for fixing it! Regards, Joe
Re: ocsp.t failing
On Sat, Dec 25, 2021 at 10:56:59AM +0100, Christophe JAILLET wrote: > Hi, > > the last travis jobs have failed on trunk because of ocsp.t. > > Unless I missed something, the only change with previous successful build is > r1896361: > > https://github.com/apache/httpd/commit/442b4b167f19e13df918402a7af28fe4a50c2730 > > I can't see any potential link. I wonder if this is due to change in the OpenSSL version in Ubuntu? The OpenSSL 3 builds have continued passing consistently. Can anybody reproduce this on an Ubuntu system? I don't see a failure on Fedora 35 with OpenSSL 1.1.1l. > BTW, ML don't receive anymore travis failures? I filed this: https://issues.apache.org/jira/browse/INFRA-22619 but not sure how to move it forward, infra have marked it waiting on user. Regards, Joe
ocsp.t failing
Hi, the last travis jobs have failed on trunk because of ocsp.t. Unless I missed something, the only change with previous successful build is r1896361: https://github.com/apache/httpd/commit/442b4b167f19e13df918402a7af28fe4a50c2730 I can't see any potential link. Any idea? BTW, ML don't receive anymore travis failures? CJ