Setting
xxx-jobset-repeats = patchelf:master:2
will cause Hydra to perform every build step in the specified jobset 2
additional times (i.e. 3 times in total). Non-determinism is not fatal
unless the derivation has the attribute "isDeterministic = true"; we
just note the lack of determinism in the Hydra database. This will
allow us to get stats about the (lack of) reproducibility of all of
Nixpkgs.
Builds can now specify the attribute "isDeterministic = true" to tell
Hydra to build with build-repeat > 0. If there is a mismatch between
rounds, the step / build fails with a suitable status.
Maybe this should be a meta attribute, but that makes it invisible to
hydra-queue-runner, and it seems reasonable to make a claim of
mandatory determinism part of the derivation (since e.g. enabling this
flag should trigger a rebuild).
We now take into account the memory necessary for compressing the NAR
being exported to the binary cache, plus xz compression overhead.
Also, we now release the memory tokens for the NAR accessor *after*
releasing the NAR accessor. Previously the memory for the NAR accessor
might still be in use while another thread does an allocation, causing
the maximum to be exceeded temporarily.
Also, use notify_all instead of notify_one to wake up memory token
waiters. This is not very nice, but not every waiter is requesting the
same number of tokens, so some might be able to proceed.
If a step is cancelled just as its builder step is starting,
doBuildStep() will return sRetry. This causes builder() to make the
step runnable again, since the queue monitor may have added new builds
referencing it. The idea is that if the latter condition is not true,
the step's reference count will drop to zero and it will be
deleted. However, if the dispatcher thread sees and locks the step
before the reference count can drop to zero in the builder thread, the
dispatcher thread will start a new builder thread for the step. Thus
the step can be kept alive for an indefinite amount of time.
The fix is for State::builder() to use a weak pointer to the step, to
ensure that the step's reference count can drop to zero *before* it's
added to the runnable queue.
This was a bad idea because pthread_cancel() is unsalvageable broken
in C++. Destructors are not allowed to throw exceptions (especially in
C++11), but pthread_cancel() can cause a __cxxabiv1::__forced_unwind
exception inside any destructor that invokes a cancellation
point. (This exception can be caught but *must* be rethrown.) So let's
just kill the builder process instead.
It was hitting
assert(reservation.unique());
Since we do want the machine reservation to be released before calling
wakeDispatcher(), let's use a different object for keeping track of
active steps.
We now kill active build steps when there are no more referring
builds. This is useful e.g. for preventing cancelled multi-hour TPC-H
benchmark runs from hogging build machines.
If two active steps of the same build failed, then the first would be
marked as "failed", but the second would end up as "orphaned", causing
it to be marked as "aborted" later on. Now it's correctly marked as
"failed".
‘basicDrv.inputSrcs’ also contains the outputs of inputDrvs. These
don't necessarily exist in the local store, so copying them may cause
an exception. We should only copy the real inputSrcs.
This rewrites the top-level loop of hydra-evaluator in C++. The Perl
stuff is moved into hydra-eval-jobset. (Rewriting the entire evaluator
would be nice but is a bit too much work.) The new version has some
advantages:
* It can run multiple jobset evaluations in parallel.
* It uses PostgreSQL notifications so it doesn't have to poll the
database. So if a jobset is triggered via the web interface or from
a GitHub / Bitbucket webhook, evaluation of the jobset will start
almost instantaneously (assuming the evaluator is not at its
concurrency limit).
* It imposes a timeout on evaluations. So if e.g. hydra-eval-jobset
hangs connecting to a Mercurial server, it will eventually be
killed.
Currently, the hydra.nixos.org queue contains 1000s of Darwin builds
that all depend on a stdenv-darwin that previously failed. However,
before, first createStep() would construct a dependency graph for each
build, then getQueuedBuilds() would discover that one of the steps had
failed previously and discard all those steps. Since the graph
construction involves a lot of uncached calls to isValidPath(), this
took several seconds per build.
Now createStep() detects the previous failure right away and bails
out.
These are build steps that remain "busy" in the database even though
they have finished, because they couldn't be updated (e.g. due to a
PostgreSQL connection problem). To prevent them from showing up as
busy in the "Machine status" page, we now periodically purge them.
Previously, if the queue monitor thread encounters a build that Hydra
has previously built, it downloaded the output paths from the binary
cache, just to determine the build products and metrics. This is very
inefficient. In particular, when doing something like merging
nixpkgs:staging into nixpkgs:master, the queue monitor thread will be
locked up for a long time fetching files from S3, causing the build
farm to be mostly idle.
Of course this is entirely unnecessary, since the build
products/metrics are already in the Hydra database. So now we just
look up a previous build with the same output path, and copy the
products/metrics.
The maximum output size per build step (as the sum of the NARs of each
output) can be set via hydra.conf, e.g.
max-output-size = 1000000000
The default is 2 GiB.
Also refactored the build error / status handling a bit.
When using a binary cache store, the queue runner receives NARs from
the build machines, compresses them, and uploads them to the
cache. However, keeping multiple large NARs in memory can cause the
queue runner to run out of memory. This can happen for instance when
it's processing multiple ISO images concurrently.
The fix is to use a TokenServer to prevent the builder threads to
store more than a certain total size of NARs concurrently (at the
moment, this is hard-coded at 4 GiB). Builder threads that cause the
limit to be exceeded will block until other threads have finished.
The 4 GiB limit does not include certain other allocations, such as
for xz compression or for FSAccessor::readFile(). But since these are
unlikely to be more than the size of the NARs and hydra.nixos.org has
32 GiB RAM, it should be fine.
Same problem as d744362e4a.
at /nix/store/ksvsbr7pg4z69bv6fbbc8h7x7rm2104m-gcc-4.9.3/include/c++/4.9.3/bits/predefined_ops.h:166
__last@entry=..., __comp=...) at /nix/store/ksvsbr7pg4z69bv6fbbc8h7x7rm2104m-gcc-4.9.3/include/c++/4.9.3/bits/stl_algo.h:1827
__comp=...) at /nix/store/ksvsbr7pg4z69bv6fbbc8h7x7rm2104m-gcc-4.9.3/include/c++/4.9.3/bits/stl_algo.h:4717
To use the local Nix store (default):
store_mode = direct
To use a local binary cache:
store_mode = local-binary-cache
binary_cache_dir = /var/lib/hydra/binary-cache
To use an S3 bucket:
store_mode = s3-binary-cache
binary_cache_s3_bucket = my-nix-bucket
Also, respect binary_cache_{secret,public}_key_file for signing the
binary cache.
This removes the "busy", "locker" and "logfile" columns, which are no
longer used by the queue runner. The "Running builds" page now only
shows builds that have an active build step.
Previously, priority bumps could take a long time to get noticed if
getQueuedBuilds() was busy processing zillions of queue
additions. (This was made worse by the reintroduction of substitute
checking.)
This allows Hydra to use binaries from available binary caches. It
makes the queue monitor thread quite a bit slower, so if you don't
want to use binary caches, it's better to add "--option
build-use-substitutes false" to the hydra-queue-runner invocation.
Fixed#243.
They will show up in machineTypes as (e.g.) x86_64-linux:local instead
of x86_64-linux. This is to prevent the Hydra provisioner from
creating machines for steps that are supposed to be executed locally.
It's easier for the Hydra provisioner to put host public keys in the
machines file than to separately manage the known_hosts file
(especially when the provisioner runs on a different machine).
This is necessary because the required system type can become
available later (e.g. by being provisioned by the
auto-scaler). However, in the future, we may want to fail steps if
they have been unsupported for more than a certain amount of time.
For example, steps that require the "kvm" feature may require a
different kind of machine to be provisioned. This can also be used to
require performance-sensitive tests to run on a particular kind of
machine, e.g., by setting requiredSystemFeatures to something like
"ec2-i2.8xlarge".
"hydra-queue-runner --status" now prints how many runnable and running
build steps exist for each machine type. This allows additional
machines to be provisioned based on the Hydra load.
Builds can now emit metrics that Hydra will store in its database and
render as time series via flot charts. Typical applications are to
keep track of performance indicators, coverage percentages, artifact
sizes, and so on.
For example, a coverage build can emit the coverage percentage as
follows:
echo "lineCoverage $pct %" > $out/nix-support/hydra-metrics
Graphs of all metrics for a job can be seen at
http://.../job/<project>/<jobset>/<job>#tabs-charts
Specific metrics are also visible at
http://.../job/<project>/<jobset>/<job>/metric/<metric>
The latter URL also allows getting the data in JSON format (e.g. via
"curl -H 'Accept: application/json'").
This prevents a race where multiple threads see that machine X is
missing path P, and start sending it concurrently. Nix handles this
correctly, but it's still wasteful (especially for the case where P ==
GHC).
A more refined scheme would be to have per machine, per path locks.
Derivations with "preferLocalBuild = true" can now be executed on
specific machines (typically localhost) by setting the mandary system
features field to include "local". For example:
localhost x86_64-linux,i686-linux - 10 100 - local
says that "localhost" can *only* do builds with "preferLocalBuild =
true". The speed factor of 100 will make the machine almost always win
over other machines.
Otherwise we never recover from reset daemon connections, e.g.
hydra-queue-runner[16106]: while loading build 599369: cannot start daemon worker: reading from file: Connection reset by peer
hydra-queue-runner[16106]: while loading build 599236: writing to file: Broken pipe
...
The error is now handled queueMonitor(), causing the next call to
queueMonitorLoop() to create a new connection.
