Re: Update strategy and timing

On 5/13/26 05:50, Pat Maddox wrote:
> On Fri, May 8, 2026, at 8:48 AM, bob prohaska wrote:
>> Is there a preferred strategy to timing updates
>> for self-hosted FreeBSD systems? 
>>
>> On the stable branches it's easy; just update when
>> updates are announced and build/install. Once caught
>> up, things can be left alone for days at least..
>>
>> With -current there's essentially no pause in the
>> stream of fresh commits, so git finds a new commit
>> by the time buildworld finishes.
>>
>> Is there some marker or indicator that signals the
>> -current tree is at least nominally consistent and
>> buildable? I'm not asking if it'll work, just whenter
>> it's worth a try.
> 
> I'm also interested in this question. One resource that hasn't been mentioned up to this point is https://ci.freebsd.org/
> 
> I don't know much about it, so I'm just inferring things based on the job names and activity log. For example, https://ci.freebsd.org/job/FreeBSD-main-amd64-test/ looks like the job that I would care about the most. Presumably you can take the commit from https://ci.freebsd.org/job/FreeBSD-main-amd64-test/lastSuccessfulBuild/parameters/ as representing the most recent commit that successfully built and passed the test suite.

"lastSuccessfulBuild" need not have passed all the tests and may not
have yet finished testing. Unsuccessful builds are not tested (likely
impossible).

For, say, https://ci.freebsd.org/job/FreeBSD-main-amd64-test/ you want
to scroll down past the pending tests (if any) and past the running test
(if any), to the completed tests. An example completed one is:

https://ci.freebsd.org/job/FreeBSD-main-amd64-test/28476/

(That is where the link takes you to.) That page shows the specific
build that was tested:

Started by upstream project FreeBSD-main-amd64-build build number 35345

FreeBSD-main-amd64-build and 35345 are links, the 35345 being to the
specific build.

It also has a link for Tests and for Show all failed tests.

> 
> One thing that's a bit confusing is that as I write this, build #28475 is listed as both the most recent successful _and_ unsuccessful build, which I don't understand.
> 
> The jobs are defined in https://github.com/freebsd/freebsd-ci/tree/main/jobs
> 
> Pat
> 
> 

Just focusing on main and overall successes vs. not start:

The most recent FreeBSD-main-amd64-test   success was on: 2026-Apr-13
The most recent FreeBSD-main-aarch64-test success was on: Never
The most recent FreeBSD-main-arm7-test    success was on: 2020-May-03
(I'll not list powerpc* ones.)

It is normal for some tests to be failing from what I can see.

By contrast, FreeBSD-*-testvm seem to be passing in recent times.
*-testvm runs before *-test and *-test only happens if *-testvm was
successful.

Only FreeBSD-main-amd64-test_* tests exist (ltp, tcptestsuite, zfs).

Avoiding failed builds and successfull builds that have FreeBSD-*-testvm
failures seems certain. But for FreeBSD-*-test it looks like judgments
about what failures are acceptable need to be involved. Then the
specific build number's link can to be used to get to the build information.


-- 
===
Mark Millard
marklmi at yahoo.com