fleet-python 0.2.120__tar.gz → 0.2.122__tar.gz

{fleet_python-0.2.120/fleet_python.egg-info → fleet_python-0.2.122}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fleet-python
-Version: 0.2.120
+Version: 0.2.122
 Summary: Python SDK for Fleet environments
 Author-email: Fleet AI <nic@fleet.so>
 License: Apache-2.0

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/__init__.py

@@ -73,7 +73,7 @@ from . import env
 from . import global_client as _global_client
 from ._async import global_client as _async_global_client
 
-__version__ = "0.2.120"
+__version__ = "0.2.122"
 
 __all__ = [
     # Core classes

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/_async/__init__.py

@@ -44,7 +44,7 @@ from ..types import VerifierFunction
 from .. import env
 from . import global_client as _async_global_client
 
-__version__ = "0.2.120"
+__version__ = "0.2.122"
 
 __all__ = [
     # Core classes

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/_async/base.py

@@ -26,7 +26,7 @@ from .exceptions import (
 try:
     from .. import __version__
 except ImportError:
-    __version__ = "0.2.120"
+    __version__ = "0.2.122"
 
 logger = logging.getLogger(__name__)
 

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/base.py

@@ -27,7 +27,7 @@ from .exceptions import (
 try:
     from . import __version__
 except ImportError:
-    __version__ = "0.2.120"
+    __version__ = "0.2.122"
 
 logger = logging.getLogger(__name__)
 

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/track/daemon.py

@@ -98,7 +98,6 @@ class Daemon:
         self._confirmed_map: dict[str, str] = {}
         self._confirmed_lock = threading.Lock()
         self._manifest_dirty = False  # set when confirmed_map gains new entries
-        self._drain_lock = threading.Lock()  # serialises claim_batch across threads
 
     # ------------------------------------------------------------------ #
     # Public entry point                                                   #
@@ -273,8 +272,7 @@ class Daemon:
         if not self._pool:
             return
 
-        with self._drain_lock:
-            items = self._queue.claim_batch(n=32)
+        items = self._queue.claim_batch(n=32)
         if not items:
             return
 

fleet_python-0.2.122/fleet/track/queue.py

@@ -0,0 +1,206 @@
+"""WAL-backed SQLite upload queue.
+
+Survives daemon crashes. Items transition: pending → in_flight → done/failed.
+Failed items are retried with exponential backoff.
+"""
+
+from __future__ import annotations
+
+import logging
+import sqlite3
+import threading
+import time
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Optional
+
+from .merkle import STATE_DB, TRACK_DIR
+
+log = logging.getLogger("fleet.track.queue")
+
+MAX_ATTEMPTS = 10
+BASE_BACKOFF_SECS = 0.5
+MAX_BACKOFF_SECS = 1800  # 30 min
+
+
+def _backoff(attempts: int) -> float:
+    return min(BASE_BACKOFF_SECS * (2 ** attempts), MAX_BACKOFF_SECS)
+
+
+@dataclass
+class QueueItem:
+    path: str        # relative to $HOME
+    sha256: str
+    attempts: int
+    last_error: Optional[str]
+
+
+class UploadQueue:
+    def __init__(self) -> None:
+        TRACK_DIR.mkdir(parents=True, exist_ok=True)
+        self._lock = threading.Lock()
+        self._conn = self._open_conn()
+
+    def _open_conn(self) -> sqlite3.Connection:
+        conn = sqlite3.connect(STATE_DB, timeout=10, check_same_thread=False)
+        conn.execute("PRAGMA journal_mode=WAL")
+        conn.execute("PRAGMA synchronous=NORMAL")
+        conn.execute("""
+            CREATE TABLE IF NOT EXISTS queue (
+                path TEXT NOT NULL,
+                sha256 TEXT NOT NULL,
+                status TEXT NOT NULL DEFAULT 'pending',
+                attempts INTEGER DEFAULT 0,
+                last_error TEXT,
+                next_attempt_at INTEGER NOT NULL DEFAULT 0,
+                enqueued_at INTEGER NOT NULL,
+                updated_at INTEGER NOT NULL,
+                UNIQUE(path, sha256)
+            )
+        """)
+        conn.execute("CREATE INDEX IF NOT EXISTS queue_status ON queue(status, next_attempt_at)")
+        conn.commit()
+
+        result = conn.execute("PRAGMA integrity_check").fetchone()
+        if result[0] != "ok":
+            conn.close()
+            self._wipe_and_reinit()
+            return self._open_conn()
+
+        return conn
+
+    def _wipe_and_reinit(self) -> None:
+        """Delete corrupt database files and start fresh. Queue items are lost but
+        the next reconcile will re-enqueue anything not yet on S3."""
+        log.warning("queue database is corrupt — wiping and reinitialising (next reconcile will re-upload missing files)")
+        for suffix in ("", "-shm", "-wal"):
+            p = Path(str(STATE_DB) + suffix)
+            if p.exists():
+                p.unlink(missing_ok=True)
+
+    def enqueue(self, path: str, sha256: str) -> None:
+        """Add path to the queue. Idempotent: same (path, sha256) is a no-op."""
+        now = int(time.time())
+        with self._lock:
+            self._conn.execute(
+                """
+                INSERT OR IGNORE INTO queue (path, sha256, status, next_attempt_at, enqueued_at, updated_at)
+                VALUES (?, ?, 'pending', 0, ?, ?)
+                """,
+                (path, sha256, now, now),
+            )
+            self._conn.commit()
+
+    def enqueue_batch(self, items: list[tuple[str, str]]) -> None:
+        now = int(time.time())
+        with self._lock:
+            self._conn.executemany(
+                """
+                INSERT OR IGNORE INTO queue (path, sha256, status, next_attempt_at, enqueued_at, updated_at)
+                VALUES (?, ?, 'pending', 0, ?, ?)
+                """,
+                [(path, sha256, now, now) for path, sha256 in items],
+            )
+            self._conn.commit()
+
+    def claim_batch(self, n: int = 16) -> list[QueueItem]:
+        """Atomically claim up to n pending items for upload."""
+        now = int(time.time())
+        with self._lock:
+            rows = self._conn.execute(
+                """
+                SELECT path, sha256, attempts, last_error FROM queue
+                WHERE status = 'pending' AND next_attempt_at <= ?
+                ORDER BY enqueued_at ASC
+                LIMIT ?
+                """,
+                (now, n),
+            ).fetchall()
+
+            if not rows:
+                return []
+
+            # Match on (path, sha256) — the natural key — so we never transition a
+            # row that wasn't in the SELECT result. Two rows can share a path with
+            # different sha256 values; path-only matching would orphan the extra row
+            # (set to in_flight with no upload attempted and no callback to resolve it).
+            placeholders = ",".join("(?,?)" for _ in rows)
+            pairs = [v for r in rows for v in (r[0], r[1])]
+            self._conn.execute(
+                f"UPDATE queue SET status = 'in_flight', updated_at = ? WHERE status = 'pending' AND (path, sha256) IN ({placeholders})",
+                [now] + pairs,
+            )
+            self._conn.commit()
+
+        return [QueueItem(r[0], r[1], r[2], r[3]) for r in rows]
+
+    def mark_done(self, path: str, sha256: str) -> None:
+        now = int(time.time())
+        with self._lock:
+            self._conn.execute(
+                "UPDATE queue SET status = 'done', updated_at = ? WHERE path = ? AND sha256 = ?",
+                (now, path, sha256),
+            )
+            # Prune rows for the same path that were enqueued *before* this one —
+            # they are older versions superseded by this upload.
+            # Rows enqueued *after* (newer file content) are preserved so they
+            # upload without waiting for the next reconcile.
+            self._conn.execute(
+                """
+                DELETE FROM queue
+                WHERE path = ? AND sha256 != ? AND status IN ('failed', 'pending')
+                  AND enqueued_at <= (
+                      SELECT enqueued_at FROM queue WHERE path = ? AND sha256 = ?
+                  )
+                """,
+                (path, sha256, path, sha256),
+            )
+            self._conn.commit()
+
+    def mark_failed(self, path: str, sha256: str, error: str) -> None:
+        now = int(time.time())
+        with self._lock:
+            row = self._conn.execute(
+                "SELECT attempts FROM queue WHERE path = ? AND sha256 = ?", (path, sha256)
+            ).fetchone()
+            attempts = (row[0] if row else 0) + 1
+            status = "failed" if attempts >= MAX_ATTEMPTS else "pending"
+            next_attempt = now + int(_backoff(attempts))
+            self._conn.execute(
+                """
+                UPDATE queue
+                SET status = ?, attempts = ?, last_error = ?, next_attempt_at = ?, updated_at = ?
+                WHERE path = ? AND sha256 = ?
+                """,
+                (status, attempts, error[:500], next_attempt, now, path, sha256),
+            )
+            self._conn.commit()
+
+    def reset_failed(self) -> int:
+        """Re-queue permanently failed items (used by reconciliation loop)."""
+        now = int(time.time())
+        with self._lock:
+            cur = self._conn.execute(
+                "UPDATE queue SET status = 'pending', attempts = 0, next_attempt_at = 0, updated_at = ? WHERE status = 'failed'",
+                (now,),
+            )
+            self._conn.commit()
+        return cur.rowcount
+
+    def remove_done(self) -> None:
+        """Purge successfully uploaded items older than 24h."""
+        cutoff = int(time.time()) - 86400
+        with self._lock:
+            self._conn.execute("DELETE FROM queue WHERE status = 'done' AND updated_at < ?", (cutoff,))
+            self._conn.commit()
+
+    def stats(self) -> dict[str, int]:
+        with self._lock:
+            rows = self._conn.execute(
+                "SELECT status, COUNT(*) FROM queue GROUP BY status"
+            ).fetchall()
+        return {row[0]: row[1] for row in rows}
+
+    def close(self) -> None:
+        with self._lock:
+            self._conn.close()

{fleet_python-0.2.120 → fleet_python-0.2.122}/fleet/track/uploader.py

@@ -19,7 +19,9 @@ from .scrubber import scrub_bytes
 log = logging.getLogger("fleet.track.uploader")
 
 WORKERS = 8
-UPLOAD_TIMEOUT_SECS = 120
+# Per-phase timeouts: connect/read can be short, write needs to be generous
+# for large JSONL files on slow connections.
+UPLOAD_TIMEOUT = httpx.Timeout(connect=10.0, read=30.0, write=300.0, pool=10.0)
 
 
 def _read_safe(path: Path) -> Optional[bytes]:
@@ -55,7 +57,7 @@ def upload_one(path: Path, presigned_url: str) -> bool:
             presigned_url,
             content=scrubbed,
             headers={"Content-Type": "application/octet-stream"},
-            timeout=UPLOAD_TIMEOUT_SECS,
+            timeout=UPLOAD_TIMEOUT,
         )
         if resp.status_code not in (200, 204):
             log.warning("upload %s → HTTP %s", path.name, resp.status_code)

{fleet_python-0.2.120 → fleet_python-0.2.122/fleet_python.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fleet-python
-Version: 0.2.120
+Version: 0.2.122
 Summary: Python SDK for Fleet environments
 Author-email: Fleet AI <nic@fleet.so>
 License: Apache-2.0

{fleet_python-0.2.120 → fleet_python-0.2.122}/pyproject.toml

@@ -5,7 +5,7 @@ build-backend = "setuptools.build_meta"
 [project]
 name = "fleet-python"
 
-version = "0.2.120"
+version = "0.2.122"
 description = "Python SDK for Fleet environments"
 authors = [
     {name = "Fleet AI", email = "nic@fleet.so"},

fleet_python-0.2.120/fleet/track/queue.py

@@ -1,158 +0,0 @@
-"""WAL-backed SQLite upload queue.
-
-Survives daemon crashes. Items transition: pending → in_flight → done/failed.
-Failed items are retried with exponential backoff.
-"""
-
-from __future__ import annotations
-
-import sqlite3
-import time
-from dataclasses import dataclass
-from pathlib import Path
-from typing import Optional
-
-from .merkle import STATE_DB, TRACK_DIR
-
-MAX_ATTEMPTS = 10
-BASE_BACKOFF_SECS = 0.5
-MAX_BACKOFF_SECS = 1800  # 30 min
-
-
-def _backoff(attempts: int) -> float:
-    return min(BASE_BACKOFF_SECS * (2 ** attempts), MAX_BACKOFF_SECS)
-
-
-@dataclass
-class QueueItem:
-    path: str        # relative to $HOME
-    sha256: str
-    attempts: int
-    last_error: Optional[str]
-
-
-class UploadQueue:
-    def __init__(self) -> None:
-        TRACK_DIR.mkdir(parents=True, exist_ok=True)
-        self._conn = sqlite3.connect(STATE_DB, timeout=10, check_same_thread=False)
-        self._conn.execute("PRAGMA journal_mode=WAL")
-        self._conn.execute("PRAGMA synchronous=NORMAL")
-        self._conn.execute("""
-            CREATE TABLE IF NOT EXISTS queue (
-                path TEXT NOT NULL,
-                sha256 TEXT NOT NULL,
-                status TEXT NOT NULL DEFAULT 'pending',
-                attempts INTEGER DEFAULT 0,
-                last_error TEXT,
-                next_attempt_at INTEGER NOT NULL DEFAULT 0,
-                enqueued_at INTEGER NOT NULL,
-                updated_at INTEGER NOT NULL,
-                UNIQUE(path, sha256)
-            )
-        """)
-        self._conn.execute("CREATE INDEX IF NOT EXISTS queue_status ON queue(status, next_attempt_at)")
-        self._conn.commit()
-
-    def enqueue(self, path: str, sha256: str) -> None:
-        """Add path to the queue. Idempotent: same (path, sha256) is a no-op."""
-        now = int(time.time())
-        self._conn.execute(
-            """
-            INSERT OR IGNORE INTO queue (path, sha256, status, next_attempt_at, enqueued_at, updated_at)
-            VALUES (?, ?, 'pending', 0, ?, ?)
-            """,
-            (path, sha256, now, now),
-        )
-        self._conn.commit()
-
-    def enqueue_batch(self, items: list[tuple[str, str]]) -> None:
-        now = int(time.time())
-        self._conn.executemany(
-            """
-            INSERT OR IGNORE INTO queue (path, sha256, status, next_attempt_at, enqueued_at, updated_at)
-            VALUES (?, ?, 'pending', 0, ?, ?)
-            """,
-            [(path, sha256, now, now) for path, sha256 in items],
-        )
-        self._conn.commit()
-
-    def claim_batch(self, n: int = 16) -> list[QueueItem]:
-        """Atomically claim up to n pending items for upload."""
-        now = int(time.time())
-        rows = self._conn.execute(
-            """
-            SELECT path, sha256, attempts, last_error FROM queue
-            WHERE status = 'pending' AND next_attempt_at <= ?
-            ORDER BY enqueued_at ASC
-            LIMIT ?
-            """,
-            (now, n),
-        ).fetchall()
-
-        if not rows:
-            return []
-
-        # Match on (path, sha256) — the natural key — so we never transition a
-        # row that wasn't in the SELECT result. Two rows can share a path with
-        # different sha256 values; path-only matching would orphan the extra row
-        # (set to in_flight with no upload attempted and no callback to resolve it).
-        placeholders = ",".join("(?,?)" for _ in rows)
-        pairs = [v for r in rows for v in (r[0], r[1])]
-        self._conn.execute(
-            f"UPDATE queue SET status = 'in_flight', updated_at = ? WHERE status = 'pending' AND (path, sha256) IN ({placeholders})",
-            [now] + pairs,
-        )
-        self._conn.commit()
-
-        return [QueueItem(r[0], r[1], r[2], r[3]) for r in rows]
-
-    def mark_done(self, path: str, sha256: str) -> None:
-        now = int(time.time())
-        self._conn.execute(
-            "UPDATE queue SET status = 'done', updated_at = ? WHERE path = ? AND sha256 = ?",
-            (now, path, sha256),
-        )
-        self._conn.commit()
-
-    def mark_failed(self, path: str, sha256: str, error: str) -> None:
-        now = int(time.time())
-        row = self._conn.execute(
-            "SELECT attempts FROM queue WHERE path = ? AND sha256 = ?", (path, sha256)
-        ).fetchone()
-        attempts = (row[0] if row else 0) + 1
-        status = "failed" if attempts >= MAX_ATTEMPTS else "pending"
-        next_attempt = now + int(_backoff(attempts))
-        self._conn.execute(
-            """
-            UPDATE queue
-            SET status = ?, attempts = ?, last_error = ?, next_attempt_at = ?, updated_at = ?
-            WHERE path = ? AND sha256 = ?
-            """,
-            (status, attempts, error[:500], next_attempt, now, path, sha256),
-        )
-        self._conn.commit()
-
-    def reset_failed(self) -> int:
-        """Re-queue permanently failed items (used by reconciliation loop)."""
-        now = int(time.time())
-        cur = self._conn.execute(
-            "UPDATE queue SET status = 'pending', attempts = 0, next_attempt_at = 0, updated_at = ? WHERE status = 'failed'",
-            (now,),
-        )
-        self._conn.commit()
-        return cur.rowcount
-
-    def remove_done(self) -> None:
-        """Purge successfully uploaded items older than 24h."""
-        cutoff = int(time.time()) - 86400
-        self._conn.execute("DELETE FROM queue WHERE status = 'done' AND updated_at < ?", (cutoff,))
-        self._conn.commit()
-
-    def stats(self) -> dict[str, int]:
-        rows = self._conn.execute(
-            "SELECT status, COUNT(*) FROM queue GROUP BY status"
-        ).fetchall()
-        return {row[0]: row[1] for row in rows}
-
-    def close(self) -> None:
-        self._conn.close()