Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Three instances send commands and DNS queries to each other.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Control protocol uses \n as command delimiter so we cannot allow \n
inside commands. Users should be able to Lua-escape \n with \010.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

For now we still attempt to support systems without lua-cqueues,
e.g. Ubuntu 16.04.

lua-http depends on lua-cqueues so the test for availability of lua-http
at the beginning of kr_https_fetch() should cover lua-cqueues as well.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

We were exceeding CI time limit while running config tests under
Valgrind so let's see if smaller number of tests will get us through CI.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

map() command on leader instance now:
- detects call errors on followers
- detects unsupported number of return values
- detects unsupported data types which cannot be serialized
- keeps nil return values (signaled by table counter "n")

Fixes: #662

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Let's detect syntax errors before sending the command to all instances.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Helper functions for handling variable length arrays with nil values.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Reasons:
- It allows to run several kresd instances and use map() without
  worrying about instance parameters.
- It removes special case from map().

Proper functionality depends on changes in MR !1011.

Petr Špaček authored 4 years ago and Tomas Krizek committed 4 years ago

Serializes: boolean, nil, number, string, table.
Skips all other types (functions, cdata, thread ...) and repeated
references to tables.

Resulting string should Lua-evaluate to identical objects.

Lukas Jezek authored 4 years ago and Tomas Krizek committed 4 years ago

This change allows map() to work with systemd integration.

As a bonus the new client implementation is based on Lua cqueues
allows caller to wrap map() in worker.corroutine() and get
asynchronous execution/avoid blocking main loop.

Currently socket communication does not employ timeouts so a hang
instance will lead to hang map() call. This does not affect query
processing _if_ map() is being run in worker.corroutine.

Fixes: #554
Fixes: #620

FIXME: see FIXMEs in diff, document the API change, re-review.

Tomas Krizek authored 4 years ago and Petr Špaček committed 4 years ago

Using exit code 77 lets meson know to classify tests as skipped instead
of failed or passed.

Works since meson 0.37.

https://mesonbuild.com/Unit-tests.html#skipped-tests-and-hard-errors

Vladimír Čunát authored 4 years ago and Petr Špaček committed 4 years ago

See the definition of kr_cdb_pt for details.

(Tiny nitpicks addressed by vcunat.)

Vladimír Čunát authored 4 years ago and Petr Špaček committed 4 years ago

... in case of usage from kresd (GC does it a bit differently).

It is very useful when debugging. This code gets executed only with
special DEBUG policy so we do not need to worry about maximum performance.

In particular this gets rid of last light user data inside kresd.

It was still causing problems on some systems, for example Debian Sid.
The error was the same: "bad light userdata pointer" from luajit,
but note that the problem can still be triggered by lua libraries,
e.g. cqueues.

It was only generating noise in test logs, especially when network is
not abvailable/is intentionally disabled.

Vladimír Čunát authored 4 years ago and Petr Špaček committed 4 years ago

Attacker might generate fake NS records pointing to victim's DNS zone.
If the zone contains wildcard the attacker might force us into packet
exchange with a (lame) DNS server on that IP address.

We now limit number of consecuctive failures and kill whole request if
limit is exceeded.

Vladimír Čunát authored 5 years ago and Petr Špaček committed 4 years ago

CWE-406: Insufficient Control of Network Message Volume (Network Amplification)

We now limit number of failed NS name resolution attempts for each
request. This does not prevent attacker from spoofing delegations
but it puts upper bound on amplification factor.

Now it works again with the latest gdb-9.1.
As a side effect, some simplification was possible, so that some
typedefs are newly defined at once with the underlying type.

Otherwise people could get confusing errors like:
> attempt to index field 'bg_worker' (a nil value)

Some rules need it and it was nil until now.

The new allocation approach isn't perfectly optimal, but it seems
relatively easy to understand and handles OOM conditions OK (I think).