Describe the bug

nix crashes when rebuilding my flake based configuration. This is the second time this specific errors to me but it is not deterministic as rerunning the same command succeeds afterwards.

some context

❯ sudo coredumpctl debug                                                                                                                                                              ~/nixpkgs
           PID: 2085 (nix)
           UID: 0 (root)
           GID: 0 (root)
        Signal: 6 (ABRT)
     Timestamp: Fri 2020-11-13 14:03:08 CET (1min 12s ago)
  Command Line: nix --experimental-features nix-command flakes build git+file:///home/teto/home#nixosConfigurations."jedha".config.system.build.toplevel --impure --no-write-lock-file --override-input nixpkgs-teto /home/teto/nixpkgs --override-input hm /home/teto/hm --override-input nova /home/teto/nova/nova-nix --out-link /tmp/nixos-rebuild.stYtk8/result
    Executable: /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/bin/nix
 Control Group: /user.slice/user-1000.slice/session-4.scope
          Unit: session-4.scope
         Slice: user-1000.slice
       Session: 4
     Owner UID: 1000 (teto)
       Boot ID: 098bc23ed9fe47e6b5fbcd2e39e68680
    Machine ID: 6ba303eb0232411fbc60369b9affc9e7
      Hostname: jedha
       Storage: /var/lib/systemd/coredump/core.nix.0.098bc23ed9fe47e6b5fbcd2e39e68680.2085.1605272588000000.lz4
       Message: Process 2085 (nix) of user 0 dumped core.

and the coredump

#0  0x00007fd6cb5c641a in raise () from /nix/store/kah5n342wz4i0s9lz9ka4bgz91xa2i94-glibc-2.32/lib/libc.so.6
#1  0x00007fd6cb5b0523 in abort () from /nix/store/kah5n342wz4i0s9lz9ka4bgz91xa2i94-glibc-2.32/lib/libc.so.6
#2  0x00007fd6cb5b041f in __assert_fail_base.cold.0 () from /nix/store/kah5n342wz4i0s9lz9ka4bgz91xa2i94-glibc-2.32/lib/libc.so.6
#3  0x00007fd6cb5bed52 in __assert_fail () from /nix/store/kah5n342wz4i0s9lz9ka4bgz91xa2i94-glibc-2.32/lib/libc.so.6
#4  0x00007fd6cbd02579 in nix::DerivationGoal::registerOutputs() () from /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/lib/libnixstore.so
#5  0x00007fd6cbd03914 in nix::DerivationGoal::buildDone() () from /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/lib/libnixstore.so
#6  0x00007fd6cbd4f563 in nix::Worker::run(std::set<std::shared_ptr<nix::Goal>, nix::CompareGoalPtrs, std::allocator<std::shared_ptr<nix::Goal> > > const&) ()
   from /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/lib/libnixstore.so
#7  0x00007fd6cbd42ebb in nix::LocalStore::buildPaths(std::vector<nix::StorePathWithOutputs, std::allocator<nix::StorePathWithOutputs> > const&, nix::BuildMode) ()
   from /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/lib/libnixstore.so
#8  0x0000000000556c87 in nix::build(nix::ref<nix::Store>, nix::Realise, std::vector<std::shared_ptr<nix::Installable>, std::allocator<std::shared_ptr<nix::Installable> > >, nix::BuildMode)
    ()
#9  0x00000000004f680a in CmdBuild::run(nix::ref<nix::Store>) ()
#10 0x0000000000506638 in nix::StoreCommand::run() ()
#11 0x000000000056e1e8 in nix::mainWrapped(int, char**) ()
#12 0x00007fd6cbfe8edd in nix::handleExceptions(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::function<void ()>) ()
   from /nix/store/fb8rfml6sr1d92zyfxzz1ax75a8djjcl-nix-2.4pre20201102_550e11f/lib/libnixmain.so
#13 0x00000000004713cb in main ()

Steps To Reproduce Too many custom repos involved. I hope the backtrace is enough.

Expected behavior

Huh, the two assert(newInfo.ca) in derivation-goal.cc look good to me. Are you repairing or checking paths by any chance?

I am just running a nixos-rebuild, at some point it's possible it calls builtins.pathExists.

something I've noticed is that happens after I rebase on nixos-unstable. It rebuilds something and fails once, then I restart the command and it works fine until the next rebase.

Hmm, that does sound like something to do with fixed output derivations.

Just came across the same error while running nixops, also appears to be nondeterministic. I'm way out of my depth here but thought I'd post some of my info in case it's to any use.

> nix --version
nix (Nix) 3.0pre20201020_e0ca98c

> sudo coredumpctl debug
           PID: 15627 (nix-daemon)
           UID: 0 (root)
           GID: 0 (root)
        Signal: 6 (ABRT)
     Timestamp: Thu 2020-11-26 12:16:02 CET (2h 3min ago)
  Command Line: nix-daemon 15622
    Executable: /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/bin/nix
 Control Group: /system.slice/nix-daemon.service
          Unit: nix-daemon.service
         Slice: system.slice
       Boot ID: 9fd49a62ddd1443dbee27f2b36fc192a
    Machine ID: b2a66e78323a41eea3abac32cebf1bef
      Hostname: nixpix
       Storage: /var/lib/systemd/coredump/core.nix-daemon.0.9fd49a62ddd1443dbee27f2b36fc192a.15627.1606389362000000.lz4
       Message: Process 15627 (nix-daemon) of user 0 dumped core.

The stacktrace from gdb:

(gdb) thread apply all bt

Thread 3 (Thread 0x7fc97200c700 (LWP 21283)):
#0  0x00007fc974292dd9 in poll () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#1  0x00007fc973d31596 in Curl_poll () from /nix/store/wmp5y493sfx864nrmn4gy94gza0d30ck-curl-7.72.0/lib/libcurl.so.4
#2  0x00007fc973d26ac3 in Curl_multi_wait.part.0 () from /nix/store/wmp5y493sfx864nrmn4gy94gza0d30ck-curl-7.72.0/lib/libcurl.so.4
#3  0x00007fc973d26ca3 in curl_multi_wait () from /nix/store/wmp5y493sfx864nrmn4gy94gza0d30ck-curl-7.72.0/lib/libcurl.so.4
#4  0x00007fc9749968d1 in nix::curlFileTransfer::workerThreadMain() () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#5  0x00007fc974997e0c in nix::curlFileTransfer::workerThreadEntry() () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#6  0x00007fc9745b9930 in ?? () from /nix/store/vran8acwir59772hj4vscr7zribvp7l5-gcc-9.3.0-lib/lib/libstdc++.so.6
#7  0x00007fc97436bead in start_thread () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libpthread.so.0
#8  0x00007fc97429cd2f in clone () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6

Thread 2 (Thread 0x7fc97180b700 (LWP 15628)):
#0  0x00007fc974292dd9 in poll () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#1  0x00000000004ed6de in std::thread::_State_impl<std::thread::_Invoker<std::tuple<nix::MonitorFdHup::MonitorFdHup(int)::{lambda()#1}> > >::_M_run() ()
#2  0x00007fc9745b9930 in ?? () from /nix/store/vran8acwir59772hj4vscr7zribvp7l5-gcc-9.3.0-lib/lib/libstdc++.so.6
#3  0x00007fc97436bead in start_thread () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libpthread.so.0
#4  0x00007fc97429cd2f in clone () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6

Thread 1 (Thread 0x7fc9731c9a00 (LWP 15627)):
#0  0x00007fc9741dd08a in raise () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#1  0x00007fc9741c7528 in abort () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#2  0x00007fc9741c740f in __assert_fail_base.cold.0 () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#3  0x00007fc9741d5a02 in __assert_fail () from /nix/store/9df65igwjmf2wbw0gbrrgair6piqjgmi-glibc-2.31/lib/libc.so.6
#4  0x00007fc97490d5cc in nix::DerivationGoal::registerOutputs() () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#5  0x00007fc97490e944 in nix::DerivationGoal::buildDone() () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#6  0x00007fc97495a603 in nix::Worker::run(std::set<std::shared_ptr<nix::Goal>, nix::CompareGoalPtrs, std::allocator<std::shared_ptr<nix::Goal> > > const&) () from /nix/store/d3mli5a7ac4li
9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#7  0x00007fc97494df5b in nix::LocalStore::buildPaths(std::vector<nix::StorePathWithOutputs, std::allocator<nix::StorePathWithOutputs> > const&, nix::BuildMode) () from /nix/store/d3mli5a7
ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#8  0x00007fc97496d1d9 in nix::daemon::processConnection(nix::ref<nix::Store>, nix::FdSource&, nix::FdSink&, nix::daemon::TrustedFlag, nix::daemon::RecursiveFlag, std::function<void (nix::
Store&)>) () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixstore.so
#9  0x00000000004b4215 in std::_Function_handler<void (), daemonLoop(char**)::{lambda()#1}>::_M_invoke(std::_Any_data const&) ()
#10 0x00007fc974797cac in std::_Function_handler<void (), nix::startProcess(std::function<void ()>, nix::ProcessOptions const&)::{lambda()#1}>::_M_invoke(std::_Any_data const&) () from /ni
x/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixutil.so
#11 0x00007fc974794659 in nix::doFork(bool, std::function<void ()>) () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixutil.so
#12 0x00007fc974797b99 in nix::startProcess(std::function<void ()>, nix::ProcessOptions const&) () from /nix/store/d3mli5a7ac4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixuti
l.so
#13 0x00000000004b53d3 in main_nix_daemon(int, char**) ()
#14 0x000000000056b29b in nix::mainWrapped(int, char**) ()
#15 0x00007fc974bf3ddd in nix::handleExceptions(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, std::function<void ()>) () from /nix/store/d3mli5a7a
c4li9ixxnjxyn2dhyv1v7ws-nix-3.0pre20201020_e0ca98c/lib/libnixmain.so
#16 0x00000000004702eb in main ()

Last output from nixops:

post-installation fixup
shrinking RPATHs of ELF executables and libraries in /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0
shrinking /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0/lib/libcares.so.2.3.0
  CC      networking/udhcp/common.o
gzipping man pages under /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0/share/man/
  CC      networking/udhcp/d6_dhcpc.o
strip is /nix/store/6k3b0kpajqnzir2603vpyzi7qqhs1l31-binutils-2.31.1/bin/strip
stripping (with command strip and flags -S) in /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0/lib
patching script interpreter paths in /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0
mv -f .deps/dirsearch.Tpo .deps/dirsearch.Po
g++ -DHAVE_CONFIG_H -I.  -I. -I./../include -I./../elfcpp -DLOCALEDIR="\"/nix/store/r52p03fm2nd3xsjvlp7nrqp5a23l6qxk-armv7l-unknown-linux-musleabihf-binutils-2.31.1/share/locale\"" -DBINDI
R="\"/nix/store/r52p03fm2nd3xsjvlp7nrqp5a23l6qxk-armv7l-unknown-linux-musleabihf-binutils-2.31.1/bin\"" -DTOOLBINDIR="\"/nix/store/r52p03fm2nd3xsjvlp7nrqp5a23l6qxk-armv7l-unknown-linux-mus
leabihf-binutils-2.31.1/armv7l-unknown-linux-musleabihf/bin\"" -DTOOLLIBDIR="\"/nix/store/r52p03fm2nd3xsjvlp7nrqp5a23l6qxk-armv7l-unknown-linux-musleabihf-binutils-2.31.1/armv7l-unknown-li
nux-musleabihf/lib\""   -W -Wall    -Wstack-usage=262144 -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -frandom-seed=dynobj.o  -pthread -g -O2 -MT dynobj.o -MD -MP -MF .deps/dynobj.Tpo -c -o
dynobj.o dynobj.cc
checking for references to /build/ in /nix/store/hzgrsyvqdm7bidisfxy8z9n0jx32a1bb-c-ares-1.15.0...
error: --- EndOfFile ------------------------------------------------------------- nix-build
unexpected end-of-file
Traceback (most recent call last):
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/bin/..nixops-wrapped-wrapped", line 991, in <module>
    args.op()
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/bin/..nixops-wrapped-wrapped", line 412, in op_deploy
    max_concurrent_activate=args.max_concurrent_activate)
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/lib/python2.7/site-packages/nixops/deployment.py", line 1063, in deploy
    self.run_with_notify('deploy', lambda: self._deploy(**kwargs))
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/lib/python2.7/site-packages/nixops/deployment.py", line 1052, in run_with_notify
    f()
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/lib/python2.7/site-packages/nixops/deployment.py", line 1063, in <lambda>
    self.run_with_notify('deploy', lambda: self._deploy(**kwargs))
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/lib/python2.7/site-packages/nixops/deployment.py", line 1003, in _deploy
    self.configs_path = self.build_configs(dry_run=dry_run, repair=repair, include=include, exclude=exclude)
  File "/nix/store/nfq1hckbwhcmw1b1c6spwsppp1px0pc4-nixops-1.7/lib/python2.7/site-packages/nixops/deployment.py", line 671, in build_configs
    raise Exception("unable to build all machine configurations")
Exception: unable to build all machine configurations

The active derivation builder.sh seems to have continued to completion after the fact.

Just got the same when rebuilding with flakes:

I've also been observing this on a bunch of systems. The weird thing is, restarting the same build a second time seems to just work?

I have just reproduced this with a Nix built with debug flags, so I now have a full backtrace with line and stack information and a 1.3 GB core file. Please let me know if you want the file to investigate this further then I can see if I can upload this somewhere.

Nix was built from 5a6ddb3de14a1684af6c793d663764d093fa7846 (current version of nixFlakes in 20.09.20210206.002c001).

The assertion failure is located here: https://github.com/NixOS/nix/blob/5a6ddb3de14a1684af6c793d663764d093fa7846/src/libstore/build/derivation-goal.cc#L3214

Looks like there is a lot of more useful information here by now, but I’ll quickly jump in to say that I have just encountered the same with nix (Nix) 2.4pre20201201_5a6ddb3 while running nix profile upgrade '.*' after updating my nixos-20.09 input.

looks like a race condition... hit it again:

building '/nix/store/b41y1pfsq3lz7s5db38yq43fjigwr4gw-unit-kmsconvt-.service.drv'...
[2/107/256 built, 2/906/944 copied (7012.7/7840.1 MiB), 4.8 MiB DL] building unit-kmsconvt-.servicenix: src/libstore/build/local-derivation-goal.cc:2460: virtual void nix::LocalDerivationGoal::registerOutputs(): Assertion `newInfo.ca' failed.
/run/current-system/sw/bin/nixos-install: line 160:  2539 Aborted                 (core dumped) nix "${flakeFlags[@]}" build "$flake#$flakeAttr.config.system.build.toplevel" --store "$mountPoint" --extra-substituters "$sub" "${verbosity[@]}" "${extraBuildFlags[@]}" "${lockFlags[@]}" --out-link "$outLink"

with nix-env (Nix) 2.4pre20210326_dd77f71

We're hitting this too, it's quite annoying as it happens in CI... Using unstable nix from last week. Also using flakes as the others. I'm unsure if I saw it when using classical commands like nix-build.(via flake-compat).

This issue has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/content-addressed-nix-call-for-testers/12881/23

This issue has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/content-addressed-nix-call-for-testers/12881/25

I'm unsure if I saw it when using classical commands like nix-build.(via flake-compat).

I tried this today and I do see this. I will try downgrading to stable nix which is a shame.

Is there a minimal repro? (might be easier than digging through core dump?)

I can't offer what I have as it's for internal packages. I did notice that if you have a lot of derivations to build (in our case, 500+), it's almost guaranteed to happen early on while it barely happens if you only have a few. So it seems like some race condition as mentioned previously.

I suppose one could try to make a failing repro just by making a large amount of derivations. I haven't tried.

Can we confirm

• That this seems to only be happening with fixed output derivations?
• Whether the already-registered path according to nix path-info has a CA? (We know it is already registered because that's the condition the failing assert is under.)

I tried this today and I do see this. I will try downgrading to stable nix which is a shame.

Using stable nix (2.3.10) with the exact same nix-build command has not exhibited the problem.

Whether the already-registered path according to nix path-info has a CA? (We know it is already registered because that's the condition the failing assert is under.)

Which path would that be? Here's a sample crash that we got in CI. If I see this, should I be running nix path-info on engine-junit-tests?

$ nix -L build .#engine-junit-tests -o target/junit
serde_cbor> trying https://static.crates.io/crates/serde_cbor/serde_cbor-0.11.1.crate
serde_cbor>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
serde_cbor>                                  Dload  Upload   Total   Spent    Left  Speed
serde_cbor> 100 43947  100 43947    0     0   327k      0 --:--:-- --:--:-- --:--:--  327k
serde_cbor (post)> Signing paths /nix/store/w5m1vf0xjbaaqh7wk4nyiajnmx9w08yy-serde_cbor-0.11.1.tar.gz
serde_cbor (post)> Uploading paths /nix/store/w5m1vf0xjbaaqh7wk4nyiajnmx9w08yy-serde_cbor-0.11.1.tar.gz
signal-hook> trying https://static.crates.io/crates/signal-hook/signal-hook-0.3.7.crate
signal-hook>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
signal-hook>                                  Dload  Upload   Total   Spent    Left  Speed
signal-hook> 100 49462  100 49462    0     0   670k      0 --:--:-- --:--:-- --:--:--  670k
smol> trying https://static.crates.io/crates/smol/smol-1.2.5.crate
smol>   % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
smol>                                  Dload  Upload   Total   Spent    Left  Speed
smol> 100 43577  100 43577    0     0   746k      0 --:--:-- --:--:-- --:--:--  746k
nix: src/libstore/build/local-derivation-goal.cc:2460: virtual void nix::LocalDerivationGoal::registerOutputs(): Assertion `newInfo.ca' failed.
/usr/bin/bash: line 131:    16 Aborted                 (core dumped) nix -L build .#engine-junit-tests -o target/junit

@Fuuzetsu Hmm, well yes it is hard to tell there which derivation it is. Maybe it would be easier to first edit the assert to print the drvPath and outputPath? I guess per https://en.cppreference.com/w/cpp/error/assert the comma operator will do.

Once we have those, we answer my two questions.

@Ericson2314 "the comma operator will do" no, it won't, because the string is rendered at compile time, not at runtime, and thus doesn't contain evaluated dynamic values. The assert should be replaced by

    if(!newInfo.ca) {
        /* proper logging + print debug info */
        assert(false);
    }

@zseri oh true. But it can be hacky because this is a bug that should be fixed, not a normal error that can crop up from time to time. I would just say use gdb but that is probably more work for just answering these too questions.

The comment above the line indicated that the path is already produced by something else. Is there a way to display which 2 realizations produced it?

@zseri Not built-in to my knowledge but one can query the database once we have the path.

@Ericson2314 In my comment https://github.com/NixOS/nix/issues/4256#issuecomment-774710467 I have provided a full backtrace which contains all the information you are asking for.

You can also try building robotnix to reproduce the issue. This build tries to realize a couple of thousand fixed-output derivation, so you are pretty much guaranteed to hit the bug.

{
  description = "Test";
  inputs.robotnix.url = "github:danielfullmer/robotnix";
  outputs = { self, robotnix }: {
    robotnixConfigurations."dailydriver" = robotnix.lib.robotnixSystem { device = "crosshatch"; flavor = "vanilla"; };
    defaultPackage.x86_64-linux = self.robotnixConfigurations."dailydriver".img;
  };
}

The crash happens in the frontend, because I can reproduce with a Nix 2.3.10 daemon and Nix 2.4pre daemon.

@hmenke Thanks. So I see the path is 3lc1nbhmpjc05da9r0a1777rnkbqqh85-annotations-13.0.jar the drvPath is a field on DerivationGoal not local variable, and those evidentally aren't printed by default, but this gets us closer.

This issue has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/tweag-nix-dev-update-12/13185/2

@Ericson2314 With Nix 2.4pre20210601_5985b8b I can now build the above robotnix flake without hitting the assertion (which would previously be hit more than 20 times until all outputs were realised). While this is not definitive proof, there is a good chance that #4798 did fix this issue.

Great! And #5030 might help too.

I marked this as stale due to inactivity. → More info

I haven't seen this in a long time. From my point of view this issue can be closed.

I have also not seen this for a long time, it's probably fixed.

Fingers crossed!