Struct DataDirLock 

pub struct DataDirLock { /* private fields */ }

RAII guard. Drop releases the kernel-level flock automatically (by closing the underlying file descriptor) and removes the pidfile.

Implementations

impl DataDirLock

pub fn data_dir(&self) -> &Path

Path of the data directory this lock protects. Used by callers that need to release-then-re-acquire (e.g. ResumeAfterAbort).

impl DataDirLock

pub fn acquire(data_dir: &Path) -> Result<Self>

Acquire the writer lock on data_dir. Returns immediately with Error::LockHeld if another process holds it.

pub fn acquire_or_break_stale(data_dir: &Path) -> Result<Self>

Like Self::acquire, but if the lock appears stale (pidfile names a PID that’s no longer alive), reclaim it. Refuses to break a lock held by a live named holder.

Strategy: re-attempt Self::acquire on the existing lockfile inode. When the named holder really has died, the kernel released its flock and the second attempt succeeds. When something else is genuinely holding the flock — an inherited FD outliving the named holder, or a PID-reuse race that briefly makes the pidfile lie — the second attempt still returns LockHeld and we surface Error::StaleLockBreakRefused.

We deliberately do NOT unlink the lockfile and acquire on a fresh inode: that path can leave two processes each holding “the lock” on separate inodes if the original inode’s flock is still held, violating invariant #1 (at most one process holds the writer lock).

Trait Implementations

impl Debug for DataDirLock

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Drop for DataDirLock

fn drop(&mut self)

Executes the destructor for this type.

fn pin_drop(self: Pin<&mut Self>)

🔬This is a nightly-only experimental API. (pin_ergonomics)

Execute the destructor for this type, but different to Drop::drop, it requires self to be pinned.