On hydra.nixos.org the queue runner had child processes that were
stuck handling an exception:
Thread 1 (Thread 0x7f501f7fe640 (LWP 1413473) "bld~v54h5zkhmb3"):
#0 futex_wait (private=0, expected=2, futex_word=0x7f50c27969b0 <_rtld_local+2480>) at ../sysdeps/nptl/futex-internal.h:146
#1 __lll_lock_wait (futex=0x7f50c27969b0 <_rtld_local+2480>, private=0) at lowlevellock.c:52
#2 0x00007f50c21eaee4 in __GI___pthread_mutex_lock (mutex=0x7f50c27969b0 <_rtld_local+2480>) at ../nptl/pthread_mutex_lock.c:115
#3 0x00007f50c1854bef in __GI___dl_iterate_phdr (callback=0x7f50c190c020 <_Unwind_IteratePhdrCallback>, data=0x7f501f7fb040) at dl-iteratephdr.c:40
#4 0x00007f50c190d2d1 in _Unwind_Find_FDE () from /nix/store/65hafbsx91127farbmyyv4r5ifgjdg43-glibc-2.33-117/lib/libgcc_s.so.1
#5 0x00007f50c19099b3 in uw_frame_state_for () from /nix/store/65hafbsx91127farbmyyv4r5ifgjdg43-glibc-2.33-117/lib/libgcc_s.so.1
#6 0x00007f50c190ab90 in uw_init_context_1 () from /nix/store/65hafbsx91127farbmyyv4r5ifgjdg43-glibc-2.33-117/lib/libgcc_s.so.1
#7 0x00007f50c190b08e in _Unwind_RaiseException () from /nix/store/65hafbsx91127farbmyyv4r5ifgjdg43-glibc-2.33-117/lib/libgcc_s.so.1
#8 0x00007f50c1b02ab7 in __cxa_throw () from /nix/store/dd8swlwhpdhn6bv219562vyxhi8278hs-gcc-10.3.0-lib/lib/libstdc++.so.6
#9 0x00007f50c1d01abe in nix::parseURL (url="root@cb893012.packethost.net") at src/libutil/url.cc:53
#10 0x0000000000484f55 in extraStoreArgs (machine="root@cb893012.packethost.net") at build-remote.cc:35
#11 operator() (__closure=0x7f4fe9fe0420) at build-remote.cc:79
...
Maybe the fork happened while another thread was holding some global
stack unwinding lock
(https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71744). Anyway, since
the hanging child inherits all file descriptors to SSH clients,
shutting down remote builds (via 'child.to = -1' in
State::buildRemote()) doesn't work and 'child.pid.wait()' hangs
forever.
So let's not do any significant work between fork and exec.
Re-executing this search_related on every access turned out to
create very problematic performance. If a jobset had a lot of
error output stored in the jobset, and there were many hundreds
or thousands of active jobs, this could easily cause >1Gbps of
network traffic.
Otherwise, when the port is randomly chosen (e.g. by specifying no port,
or a port of 0), it will just show that the port is 0 and not the port
that is actually serving the metrics.
This is syntactically lighter wait, and demonstates there are no weird
dynamic lifetimes involved, just regular passing reference to callee
which it only borrows for the duration of the call.
Periodically, I have seen tests fail because of out of order queue runner behavior:
checking the queue for builds > 0...
loading build 1 (tests:basic:empty_dir)
aborting unsupported build step '...-empty-dir.drv' (type 'x86_64-linux')
marking build 1 as failed
adding new machine ‘localhost’
This patch should prevent the dispatcher from running before any machines are
made available.
This in-progress feature will run a dynamically generated set of
buildFinished hooks, which must be nested under the `runCommandHook.*`
attribute set. This implementation is not very good, with some to-dos:
1. Only run if the build succeeded
2. Verify the output is named $out and that it is an executable file
(or a symlink to a file)
3. Require the jobset itself have a flag enabling the feature, since
this feature can be a bit dangerous if various people of different
trust levels can create the jobs.
This shouldn't be possible normally, but it is possible to:
$db->resultset('RunCommandLogs')->new({ uuid => "../etc/passwd" });
if you have access to the `$db`.
Also split it out to a new div -- there are now 3 lines per
RunCommandLog -- the first saying when it started, the second saying how
long it ran for (or has been running), and the third with the buttons
for the pretty, raw, and tail versions of the log.
This also adds the `runcommandlog` object to the stash so that we can
access its uuid as well as command run in order to display more useful
and specific information on the webpage.
Using a sha1 of the command combined with the build ID is not a
particularly good or unique identifier:
* A build could fail, be restarted, and then succeed -- assuming no
configuration changes, the sha1 hash of the command as well as the build
ID will be the same. This would lead to an overwritten log file.
* Allowing user input to influence filenames is not the best of ideas.
Since breaking the filename construction out to a helper function,
Hydra::Model::DB is no longer used. Importing Hydra::Helper::Nix,
however, has the potential to break tests, so just use the functions we
need without importing the entire module.
run3 just seems to do better handling for what we want to do, and
requires less deep-reaching changes to this plugin to get it to play
nice, as IPC::Run::run would.
This uses the somewhat restrictive umask of 0027 so that people outside
the user or group cannot read the files. This also helps to inhibit
TOCTOU where someone else has a handle to our file before we chmod it
and after we close it.
In a Hydra instance I saw:
possibly transient failure building ‘/nix/store/X.drv’ on ‘localhost’:
dependency '/nix/store/Y' of '/nix/store/Y.drv' does not exist,
and substitution is disabled
This is confusing because the Hydra in question does have substitution enabled.
This instance uses:
keep-outputs = true
keep-derivations = true
and an S3 binary cache which is not configured as a substituter in the nix.conf.
It appears this instance encountered a situation where store path Y was built
and present in the binary cache, and Y.drv was GC rooted on the instance,
however Y was not on the host.
When Hydra would try to build this path locally, it would look in the binary
cache to see if it was cached:
(nix)
439 bool valid = isValidPathUncached(storePath);
440
441 if (diskCache && !valid)
442 // FIXME: handle valid = true case.
443 diskCache->upsertNarInfo(getUri(), hashPart, 0);
444
445 return valid;
Since it was cached, the store path was considered Valid.
The queue monitor would then not put this input in for substitution, because
the path is valid:
(hydra)
470 if (!destStore->isValidPath(*i.second.path(*localStore, step->drv->name, i.first))) {
471 valid = false;
472 missing.insert_or_assign(i.first, i.second);
473 }
Hydra appears to correctly handle the case of missing paths that need
to be substituted from the binary cache already, but since most
Hydra instances use `keep-outputs` *and* all paths in the binary cache
originate from that machine, it is not common for a path to be cached
and not GC rooted locally.
I'll run Hydra with this patch for a while and see if we run in to the
problem again.
A big thanks to John Ericson who helped debug this particular issue.
I'm not sure this is a good implementation as-is. It does work,
but the password gets echo'd to the screen. I tried to use IO::Prompt
but IO::Prompt really seems to want to read the password from ARGV.
Deleting jobsets first would fail because buildmetrics has an FK
to the jobset. However, the jobset / project relationship is not
marked as CASCADE.
Deleting all the builds automatically cascades to delete
buildmetrics, so deleting the relevant builds first, then deleting
the jobset solves it.
