[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-21 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-220791012 @PSUdaemon There's not any issues I'm aware of at the moment. --- If your project is set up for it, you can reply to this email and have your reply appear on GitH

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-21 Thread PSUdaemon
Github user PSUdaemon commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-220764504 Are there still outstanding issues here or is it possible to merge this now? --- If your project is set up for it, you can reply to this email and have your rep

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-20 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-220702623 Fixed --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature en

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-19 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-220376323 . --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feat

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-18 Thread oknet
Github user oknet commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r63826235 --- Diff: lib/ts/Diags.cc --- @@ -625,11 +625,14 @@ Diags::should_roll_diagslog() fflush(diags_log->m_fp); if (diags_log->roll()

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-18 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-220231207 @chaiman I see what you mean. I'll fix that in a separate ticket. --- If your project is set up for it, you can reply to this email and have your reply appear on

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-18 Thread chaiman
Github user chaiman commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r63660164 --- Diff: lib/ts/Diags.cc --- @@ -642,11 +645,14 @@ Diags::should_roll_diagslog() if (diags_log->roll()) { --- End diff -- sam

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-18 Thread chaiman
Github user chaiman commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r63659601 --- Diff: lib/ts/Diags.cc --- @@ -625,11 +625,14 @@ Diags::should_roll_diagslog() fflush(diags_log->m_fp); if (diags_log->roll

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-15 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219347020 @zwoop done --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feat

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-15 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219342137 Can you squash the commits together too? We don't want intermediary commits that are failing (they make git bisect difficult to deal with). --- If your project is

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-15 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219324620 @zwoop I've gone ahead and done that. I don't know how (or if I can) to run the CI builds so I'll just leave that to you. --- If your project is set up for it, yo

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-14 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219261334 This needs a rebase and new clang-format indentation. It fails with https://ci.trafficserver.apache.org/view/github/job/Github-Linux/67/ --- If your projec

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-13 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219194926 Can one of the admins verify this patch? Only approve PRs which have been reviewed. --- If your project is set up for it, you can reply to this email and have your

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-13 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219191875 Can one of the admins verify this patch? Only approve PRs which have been reviewed. --- If your project is set up for it, you can reply to this email and have your

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-13 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219190153 Can one of the admins verify this patch? Only approve PRs which have been reviewed. --- If your project is set up for it, you can reply to this email and have your

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-13 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-219186505 Can one of the admins verify this patch? Only approve PRs which have been reviewed. --- If your project is set up for it, you can reply to this email and have your

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-12 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218929112 Upgraded git on all the build boxes, so try again [approve ci]. --- If your project is set up for it, you can reply to this email and have your reply appear on GitH

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-12 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218905112 FreeBSD v10 build finished successfully. Details on https://ci.trafficserver.apache.org/job/Github-FreeBSD/17/ --- If your project is set up for it, you can

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-12 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218904299 Linux (CentOS7) build failed! Details on https://ci.trafficserver.apache.org/job/Github-Linux/34/ --- If your project is set up for it, you can reply to thi

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-12 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218903748 One more round of builds [approve ci]. --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your proje

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-12 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218782605 Can one of the admins verify this patch? Only approve PRs which have been reviewed. --- If your project is set up for it, you can reply to this email and have your

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-11 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218614205 Linux (CentOS7) build finished successfully. Details on https://ci.trafficserver.apache.org/job/Github-Linux/17/ --- If your project is set up for it, you c

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-11 Thread zwoop
Github user zwoop commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218612393 [approve ci] --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feat

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-05-11 Thread atsci
Github user atsci commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-218525646 Can one of the admins verify this patch? --- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your pro

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-210516157 @jpeach Thanks for the comments. I've addressed the relevant issues you brought up in the force pushed commits. I've also gone ahead and made a separate ti

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread jpeach
Github user jpeach commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-210501721 I don't see any reason to block writing tests on some proposed framework. All you really need to write a test is a hook to run it (which we have) and a way to chec

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread SolidWallOfCode
Github user SolidWallOfCode commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59890598 --- Diff: lib/ts/Diags.cc --- @@ -716,14 +722,12 @@ Diags::should_roll_outputlog() bool ret_val = false; bool need_consider_stderr

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59889953 --- Diff: lib/ts/Diags.cc --- @@ -816,29 +820,39 @@ Diags::set_stdout_output(const char *_bind_stdout) if (strcmp(_bind_stdout, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59889285 --- Diff: lib/ts/Diags.cc --- @@ -716,14 +722,12 @@ Diags::should_roll_outputlog() bool ret_val = false; bool need_consider_stderr = true;

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread danobi
Github user danobi commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59888075 --- Diff: lib/ts/Diags.cc --- @@ -816,29 +820,39 @@ Diags::set_stdout_output(const char *_bind_stdout) if (strcmp(_bind_stdout, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-15 Thread SolidWallOfCode
Github user SolidWallOfCode commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-210488274 While I think unit tests are in general a good idea, I'm not sure how that could be done in this case and not in the time frame for committing this to 6.2

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread danobi
Github user danobi commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-209691924 @jpeach Is there a preferred framework for unit testing? (ie how/where do I put the tests?) --- If your project is set up for it, you can reply to this email and

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread danobi
Github user danobi commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59644646 --- Diff: lib/ts/Diags.cc --- @@ -716,14 +722,12 @@ Diags::should_roll_outputlog() bool ret_val = false; bool need_consider_stderr = true;

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread SolidWallOfCode
Github user SolidWallOfCode commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r5964 --- Diff: lib/ts/Diags.cc --- @@ -716,14 +722,12 @@ Diags::should_roll_outputlog() bool ret_val = false; bool need_consider_stderr

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread SolidWallOfCode
Github user SolidWallOfCode commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-209664725 Yes. We looked at various approaches and decided on this one because there was already a lock around access to the file descriptor. Locking the update to

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-209663459 OK the locking makes sense to me I think. You have to hold the lock because you want to serialize WRT printing output. ``Diags::rebind_stderr`` and ``Diags

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59632436 --- Diff: lib/ts/Diags.cc --- @@ -816,29 +820,39 @@ Diags::set_stdout_output(const char *_bind_stdout) if (strcmp(_bind_stdout, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59632366 --- Diff: lib/ts/Diags.cc --- @@ -716,14 +722,12 @@ Diags::should_roll_outputlog() bool ret_val = false; bool need_consider_stderr = true;

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59632028 --- Diff: lib/ts/Diags.cc --- @@ -816,29 +820,39 @@ Diags::set_stdout_output(const char *_bind_stdout) if (strcmp(_bind_stdout, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59631832 --- Diff: lib/ts/Diags.cc --- @@ -816,29 +820,39 @@ Diags::set_stdout_output(const char *_bind_stdout) if (strcmp(_bind_stdout, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59631163 --- Diff: lib/ts/Diags.cc --- @@ -853,29 +867,39 @@ Diags::set_stderr_output(const char *_bind_stderr) if (strcmp(_bind_stderr, "") == 0)

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread jpeach
Github user jpeach commented on a diff in the pull request: https://github.com/apache/trafficserver/pull/568#discussion_r59629895 --- Diff: lib/ts/Diags.cc --- @@ -642,10 +642,10 @@ Diags::should_roll_diagslog() bool ret_val = false; log_log_trace("should_roll_di

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread SolidWallOfCode
Github user SolidWallOfCode commented on the pull request: https://github.com/apache/trafficserver/pull/568#issuecomment-209628730 Looks good to me. Dan, Jason, and I discussed this at length prior to the work and this looks in line with our concensus. --- If your project is set up f

[GitHub] trafficserver pull request: TS-4072 Diagnostic log rolling races

2016-04-13 Thread danobi
GitHub user danobi opened a pull request: https://github.com/apache/trafficserver/pull/568 TS-4072 Diagnostic log rolling races This patch prevents race conditions when a diagnostic log rolls. The patch is a little longer than I had hoped for because of error handling conditions. Y