ci: allow failure of obs:build:all

See merge request !910

This job tends to fail very often, but very frequently due to issues
with OBS itself - outside of our control. The output of the job can
still be useful to check manually, e.g. before releases.

This value seems to be the default, but it is important to have
it set explicitly, otherwise when: manual actions could be skipped

https://docs.gitlab.com/ee/ci/yaml/#whenmanual

lua: stop trying to tweak lua's GC

See merge request !201

Tomas Krizek authored 5 years ago and Petr Špaček committed 5 years ago

cherry-picked from f0ca89ac, original author Vlada Cunat

TL;DR: I believe all lua_gc() calls stemmed from misunderstanding lua
documentation, and the current settings seem potentially dangerous.

First, let me rely on lua 5.1 docs, as luajit 2 is documented to have
done only minor changes in the GC.
http://www.lua.org/manual/5.1/manual.html#lua_gc
http://wiki.luajit.org/New-Garbage-Collector#rationale

Commit 5a709411 claims to have increased the speed of GC to 400 % of
speed of allocation, but LUA_GCSETSTEPMUL is the parameter that
controls that, and that one was lowered to 99 % and later in
0ee2d1d7 even to 50 %.  Documentation explicitly says that setting
the value under 100 % may cause problems.

The default values seem perfectly sane to me and currently I can't see
any particular reason to change them.  It's 200 % relative GC speed,
and waiting for allocated size to double before starting another cycle.

I assume the resulting possibility of GC being too slow cause...

daemon/io: use SO_REUSEPORT_LB if available (FreeBSD 12.0+)

See merge request !907

Vladimír Čunát authored 5 years ago and Petr Špaček committed 5 years ago

and don't use SO_REUSEPORT on non-Linux.  (Free)BSD has a different
meaning for it, which only brings confusion - only the last instance
would be getting packets.

daemon/ tty commands: don't log unless --verbose

Closes #528

See merge request !908

It's minimalistic: no change if in interactive or --verbose mode.

systemd: add env variable SYSTEMD_INSTANCE

See merge request !906

ci: updates

See merge request !905

update NEWS, bump to 4.3.0

See merge request !904

Resolve "RRset merge operation is too slow for big RRsets"

Closes #518

See merge request !903

Vladimír Čunát authored 5 years ago and Tomas Krizek committed 5 years ago

- written relatively defensively - act OK even if the API
  isn't used in an ideal way
- CI lint:scan-build: bump the error count;
  It's only another instance of the mis-detected array_push().
- the removed stale note in modules/meson.build isn't really related

iterate: fix limit on CNAME chain length

See merge request !899

Unbound has limit 10, and practically useful numbers are way lower.

The accounting was just broken and overly messy anyway.

performance: lua-related improvements

See merge request !874

Vladimír Čunát authored 5 years ago and Petr Špaček committed 5 years ago

Basically the same as the parent commit (just much simpler).

Vladimír Čunát authored 5 years ago and Petr Špaček committed 5 years ago

When all lua modules get unloaded, this change makes the module's
contribution to QPS unmeasurable (for me), saving a few percent.
The point is to almost always return very cheaply, in particular without
creating any lua GC object (like FFI for kr_query).

Note: some checks didn't make much sense, so I improved those as well.

Vladimír Čunát authored 5 years ago and Petr Špaček committed 5 years ago

Running the full special-domain checks is relatively expensive.

I've never seen anyone use postrules.

The new way of transitioning to layer callbacks - done because of
portability (mainly to aarch64) - is a bit expensive.  This is a simple
way of recovering that cost.  Merge 603a24fc regressed speed a bit.

Packaging fixes

See merge request !895