Most SSL tests failed on 9/21/2002 Netscape nightly QA with a rather strange error: elgamal.1 Wait for Server Failed This error was not seen on the Sun nightly QA that same day. Also, Tinderbox does not report any checkins on Saturday preceding this failure. Friday's nightly QA on WinNT had all passed. This eoor was also not repeated on Sunday's and Monday's nightly QA. There also no checkins between the failure and Monday early morning's QA. Errors reported and ouput log are attached in subsequent attachments.
looks like the build did not complete, or tstclnt got deleted? tstclnt: wait_for_selfserv 128: start_selfserv 170: ssl_cov 193: ./ssl.sh 342: .: all.sh 98: not found RETRY: tstclnt -p 8443 -h elgamal -q \ -d ../client < W:/nss/nsstip/builds/20020921.1/spd06_NT4/mozilla/security/nss/tests/ssl/sslreq.txt tstclnt: wait_for_selfserv 128: start_selfserv 170: ssl_cov 193: ./ssl.sh 342: .: all.sh 98: not found
Thank you for looking at these test failures, Sonja. Sonja is right. The cause of these test failures was that 'tstclnt' could not be found. Our nss/nsstip/builds/20020921.1/spd06_NT4 build tree has already been deleted, so I can't verify this, but it's obvious from the error messages in output.log. Marked the bug invalid.
I am able to verify this as network problems now. I was in the lab today and observed sporadic network problems firsthand. Anthony and IC have confirmed that akroyd (where our NT builds are written to, and where sbstools resides and our nightly QA scripts run from) has been up and down yesterday and today. Sporadic failures also occurred in the week this failure was observed.