This was preventing destructors from running. In particular, it was
preventing the deletion of the temproot file for each worker
process. It may also have been responsible for the excessive WAL
growth on Hydra (due to the SQLite database not being closed
properly).
Apparently broken by accident in
8e9140cfde.
The thread calls poll() to wait until a HUP (or other error event)
happens on the client connection. If so, it sends SIGINT to the main
thread, which is then cleaned up normally. This is much nicer than
messing around with SIGPOLL.
‘trusted-users’ is a list of users and groups that have elevated
rights, such as the ability to specify binary caches. It defaults to
‘root’. A typical value would be ‘@wheel’ to specify all users in the
wheel group.
‘allowed-users’ is a list of users and groups that are allowed to
connect to the daemon. It defaults to ‘*’. A typical value would be
‘@users’ to specify the ‘users’ group.
When copying a large path causes the daemon to run out of memory, you
now get:
error: Nix daemon out of memory
instead of:
error: writing to file: Broken pipe
Currently, clients cannot recover from an isValidPath RPC with an
invalid path parameter because the daemon closes the connection when
that happens.
More precisely:
1. in performOp, wopIsValidPath case, ‘readStorePath’ raises an
‘Error’ exception;
2. that exception is caught by the handler in ‘processConnection’;
3. the handler determines errorAllowed == false, and thus exits after
sending the message.
This last part is fixed by calling ‘startWork’ early on, as in the patch
below.
The same reasoning could be applied to all the RPCs that take one or
more store paths as inputs, but isValidPath is, by definition, likely to
be passed invalid paths in the first place, so it’s important for this
one to allow recovery.
On a system with multiple CPUs, running Nix operations through the
daemon is significantly slower than "direct" mode:
$ NIX_REMOTE= nix-instantiate '<nixos>' -A system
real 0m0.974s
user 0m0.875s
sys 0m0.088s
$ NIX_REMOTE=daemon nix-instantiate '<nixos>' -A system
real 0m2.118s
user 0m1.463s
sys 0m0.218s
The main reason seems to be that the client and the worker get moved
to a different CPU after every call to the worker. This patch adds a
hack to lock them to the same CPU. With this, the overhead of going
through the daemon is very small:
$ NIX_REMOTE=daemon nix-instantiate '<nixos>' -A system
real 0m1.074s
user 0m0.809s
sys 0m0.098s
So if a path is not garbage solely because it's reachable from a root
due to the gc-keep-outputs or gc-keep-derivations settings, ‘nix-store
-q --roots’ now shows that root.
