It would be helpful to have timestamps added to the verbose option of rebuild_zebra.pl to see how long the rebuild has been running.
Created attachment 113773 [details] [review] Bug 27048: Add basic timekeeping to rebuild_zebra
Created attachment 113780 [details] [review] Bug 27048: Add basic timekeeping to rebuild_zebra This patch adds a start time, end time and elapsed times on the 'records exported' lines to the verbose output of the rebuild_zebra.pl script. Test plan 1/ Run rebuild_zebra.pl -a -b -v 2/ Note the new timestamps included on the verbose output 3/ Signoff
Comment on attachment 113780 [details] [review] Bug 27048: Add basic timekeeping to rebuild_zebra Review of attachment 113780 [details] [review]: ----------------------------------------------------------------- ::: misc/migration_tools/rebuild_zebra.pl @@ +848,5 @@ > + my $now_pretty = POSIX::strftime("%H:%M:%S",localtime($now)); > + my $elapsed_pretty = $d ? "[$d:$h:$m:$s]" : $h ? "[$h:$m:$s]" : $m ? "[$m:$s]" : "[$s]"; > + > + return "$now_pretty $elapsed_pretty"; > +} Why not just use DateTime->now()->strftime("%H:%M:%S")?
Created attachment 115931 [details] [review] Bug 27048: Add basic timekeeping to rebuild_zebra This patch adds a start time, end time and elapsed times on the 'records exported' lines to the verbose output of the rebuild_zebra.pl script. Test plan 1/ Run rebuild_zebra.pl -a -b -v 2/ Note the new timestamps included on the verbose output 3/ Signoff Signed-off-by: David Nind <david@davidnind.com>
(In reply to David Cook from comment #3) > Comment on attachment 113780 [details] [review] [review] > Bug 27048: Add basic timekeeping to rebuild_zebra > > Review of attachment 113780 [details] [review] [review]: > ----------------------------------------------------------------- > > ::: misc/migration_tools/rebuild_zebra.pl > @@ +848,5 @@ > > + my $now_pretty = POSIX::strftime("%H:%M:%S",localtime($now)); > > + my $elapsed_pretty = $d ? "[$d:$h:$m:$s]" : $h ? "[$h:$m:$s]" : $m ? "[$m:$s]" : "[$s]"; > > + > > + return "$now_pretty $elapsed_pretty"; > > +} > > Why not just use DateTime->now()->strftime("%H:%M:%S")? Martin, can you answer David's question please?
This works well, still wondering about comment#3 - Failing QA to get attention :)
(In reply to David Cook from comment #3) > Comment on attachment 113780 [details] [review] [review] > Bug 27048: Add basic timekeeping to rebuild_zebra > > Review of attachment 113780 [details] [review] [review]: > ----------------------------------------------------------------- > > ::: misc/migration_tools/rebuild_zebra.pl > @@ +848,5 @@ > > + my $now_pretty = POSIX::strftime("%H:%M:%S",localtime($now)); > > + my $elapsed_pretty = $d ? "[$d:$h:$m:$s]" : $h ? "[$h:$m:$s]" : $m ? "[$m:$s]" : "[$s]"; > > + > > + return "$now_pretty $elapsed_pretty"; > > +} > > Why not just use DateTime->now()->strftime("%H:%M:%S")? No major reason for using POSIX.. mostly I was avoiding more DateTime use.. but as it's a single call it's not terrible so can be swapped if needed.
I have a better answer now.. 1. Because there is already president for using POSIX in the script 2. Because if we were switching to DateTime we should use dt_from_string to localise and be consistent with the rest of Koha. 3. Because DateTime::Durations don't print nicely without importing DateTime::Format::Duration which would add another dependency for a reasonably trivial functionality. 4. Because I hoped to 'keep it simple' by just working in seconds.. which it did. I've tried to write a patch to switch to using DateTime and it ends up requiring extra dependencies, looking worse in the output and more complicated to work with :(
Created attachment 117201 [details] [review] Bug 27048: (follow-up) Use DateTime instead of POSIX This introduces additional dependancies and builds DateTime objects multiple times.. I'm not sure it's worth it.
I think I must not have paid enough attention to the patch the first time around. I thought you were just adding timestamps to the loglines. I didn't realize that you were tracking elapsed time as well. I would do it differently but I'm not the one writing the code. I'll accept the reasons for using POSIX over DateTime, so no complaint from me.
Created attachment 117418 [details] [review] Bug 27048: Add basic timekeeping to rebuild_zebra This patch adds a start time, end time and elapsed times on the 'records exported' lines to the verbose output of the rebuild_zebra.pl script. Test plan 1/ Run rebuild_zebra.pl -a -b -v 2/ Note the new timestamps included on the verbose output 3/ Signoff Signed-off-by: David Nind <david@davidnind.com> Signed-off-by: Katrin Fischer <katrin.fischer.83@web.de>
I have obsoleted the datetime patch. I had already tested this before without and found it working well.
60 seconds: exporting authority 13:06:38 [1:0] 61 seconds: exporting authority 13:05:30 [1:1] 54321 seconds: exporting authority 13:07:06 [15:5:21] It's not displayed as a duration and it's confusing IMO. It should at least display the HH, MM and SS parts on 2 digits. Why not "15 hours, 5 minutes, 21 seconds" instead? Or "15h 5m 21s"?
(In reply to Jonathan Druart from comment #13) > 60 seconds: > exporting authority 13:06:38 [1:0] > 61 seconds: > exporting authority 13:05:30 [1:1] > 54321 seconds: > exporting authority 13:07:06 [15:5:21] > > It's not displayed as a duration and it's confusing IMO. It should at least > display the HH, MM and SS parts on 2 digits. > Why not "15 hours, 5 minutes, 21 seconds" instead? Or "15h 5m 21s"? Hm, you are right. I guess my test db was too small, so the big numbers didn't make me suspicious. I think the second format you suggest would be better. Or: Duration: 00:00:02
Created attachment 117452 [details] [review] Bug 27048: (follow-up) Correction to elapsed printing
Fix the duration printing to take the form [00:00:07]
Martin, don't you think we should add a "elapsed" or "duration" to explain what it is?
Created attachment 117617 [details] [review] Bug 27048: (follow-up) Correction to elapsed printing JD Amended patch: Remove perlcritic error "$d" is declared but not used at line 839, column 5. Unused variables clutter code and make it harder to read. (Severity: 3)
Pushed to master for 21.05, thanks to everybody involved!
Nice little enhancement, I choose to backport. Pushed to 20.11.x for 20.11.04
Enhancement, not backported t0 20.05