Date: Thu, 9 Oct 2025 23:17:05 -0700 From: Mark Millard <marklmi@yahoo.com> To: Graham Perrin <grahamperrin@gmail.com> Cc: freebsd-pkg@freebsd.org Subject: Re: poudriere bulk results: no end result for a queued port (was: FYI: 15stable-amd64-quarterly has had 2 successful builds, despite not being distributed yet) Message-ID: <B8AD091A-326C-4364-A788-A9EA72D405D9@yahoo.com> In-Reply-To: <5308b7ab-f840-4f02-83a2-49a2540c3e3c@gmail.com> References: <C6B33B01-A73B-41EF-B4A7-EB3B84931853@yahoo.com> <5308b7ab-f840-4f02-83a2-49a2540c3e3c@gmail.com>
index | next in thread | previous in thread | raw e-mail
On Oct 9, 2025, at 22:30, Graham Perrin <grahamperrin@gmail.com> wrote: > On 10/10/2025 04:57, Mark Millard wrote: >> Build Status Started (UTC) Elapsed >> 9a9cbb37913d done: Tue, 07 Oct 2025 08:51:36 GMT 13:00:53 >> 24fedaeb4e97 done: Thu, 09 Oct 2025 01:03:13 GMT 11:35:38 >> >> For reference: >> >> https://pkg-status.freebsd.org/beefy21/build.html?mastername=15stable-amd64-quarterly&build=9a9cbb37913d >> https://pkg-status.freebsd.org/beefy21/build.html?mastername=15stable-amd64-quarterly&build=24fedaeb4e97 >> >> >> (Prior but larger-partial builds contributed.) > > > I'm mystified when a port is queued but then, no end result when the build is done. > > Build 24fedaeb4e97, for example, x11/kde is in none of these four sets of results: > > built > failed > skipped > ignored There are differ stages of being queued(/filtered) but there is also bad terminology reuse involved in the presentation. Initial "Queued ports" stage reports: "Showing 1 to 10 of 37,164 entries". That 37,164 appears to be the total number of potential port-packages across all the ports for 2025Q4's quarterly as of the snapshot used for 24fedaeb4e97. Later (top of page) Filtered stage: "Queued" "992" (my terminology) Note: Built + Failed + Skipped + Ignored == 373 + 152 + 121 + 346 == 992 == Filtered stage: Queued 992 So they match just fine. 37164-992==36172 port-packages will not be in any of the 4 categories above. The vast majority of the port-packages that were built overall were built in 2 prior bulk runs that failed to complete ( were partial, e3fa556c9a50 and d1f407b76f7a ). Both such failures lead to a manually initiated "sigterm:" status. All 4 builds were on beefy21. I do not see much mystery here. I will also note that the counting can end up with Remaining -1 on occasion, even when the build otherwise works normally. I do not know what contributes to that. === Mark Millard marklmi at yahoo.comhome | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?B8AD091A-326C-4364-A788-A9EA72D405D9>