carlostellocba/OnEverDrive
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: implementar Fase 2 del roadmap (Modo Replica Exacta / Mirroring)

Browse Source
This commit is contained in:
Carlos Tello
2026-08-14 21:50:00 -03:00
parent a72cf1480e
commit a1cafff90e
14 changed files with 421 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files
windows-agent/agent/config.py
+5
View File
@@ -21,6 +21,11 @@ class LocalFolderJob(BaseModel):
name: str
source_path: str
file_patterns: str = "*.bak,*.mdf"
exclude_patterns: str = ""
skip_hidden: bool = False
skip_system: bool = False
skip_readonly: bool = False
sync_deletions: bool = False
schedule_cron: str = "daily"
schedule_interval_minutes: int = 60
min_stable_seconds: int = 60
windows-agent/agent/scanner.py
+53 -6
View File
@@ -45,12 +45,28 @@ def is_file_stable(filepath: Path, min_stable_seconds: int = 60, sample_interval
except Exception:
return False
import platform
import stat
class DirectoryScanner:
"""Scans Windows source paths for files matching specific backup patterns."""
def __init__(self, source_path: str, file_patterns: str = "*.bak,*.mdf", min_stable_seconds: int = 60):
def __init__(
self,
source_path: str,
file_patterns: str = "*.bak,*.mdf",
exclude_patterns: str = "",
skip_hidden: bool = False,
skip_system: bool = False,
skip_readonly: bool = False,
min_stable_seconds: int = 60
):
self.source_path = Path(source_path).resolve()
self.patterns = [p.strip() for p in file_patterns.split(",") if p.strip()]
self.exclude_patterns = [p.strip() for p in exclude_patterns.split(",") if p.strip()]
self.skip_hidden = skip_hidden
self.skip_system = skip_system
self.skip_readonly = skip_readonly
self.min_stable_seconds = min_stable_seconds
def scan(self) -> List[Path]:
@@ -61,15 +77,20 @@ class DirectoryScanner:
matched_files: List[Path] = []
if self.source_path.is_file():
if self._matches_patterns(self.source_path.name):
matched_files.append(self.source_path)
if self._matches_patterns(self.source_path.name) and not self._matches_exclude(self.source_path.name):
if not self._should_skip_by_attributes(self.source_path):
matched_files.append(self.source_path)
return matched_files
for root, _, files in os.walk(self.source_path):
for root, dirs, files in os.walk(self.source_path):
# Prune excluded directories from search in-place
dirs[:] = [d for d in dirs if not self._matches_exclude(d)]
for file in files:
if self._matches_patterns(file):
if self._matches_patterns(file) and not self._matches_exclude(file):
full_path = Path(root) / file
matched_files.append(full_path)
if not self._should_skip_by_attributes(full_path):
matched_files.append(full_path)
return matched_files
@@ -80,3 +101,29 @@ class DirectoryScanner:
if fnmatch.fnmatch(filename.lower(), pattern.lower()):
return True
return False
def _matches_exclude(self, filename: str) -> bool:
for pattern in self.exclude_patterns:
if fnmatch.fnmatch(filename.lower(), pattern.lower()):
return True
return False
def _should_skip_by_attributes(self, filepath: Path) -> bool:
if platform.system() != "Windows":
return False
try:
st = filepath.stat()
attrs = getattr(st, "st_file_attributes", 0)
if not attrs:
return False
if self.skip_readonly and (attrs & 0x01): # FILE_ATTRIBUTE_READONLY
return True
if self.skip_hidden and (attrs & 0x02): # FILE_ATTRIBUTE_HIDDEN
return True
if self.skip_system and (attrs & 0x04): # FILE_ATTRIBUTE_SYSTEM
return True
except Exception:
pass
return False
windows-agent/agent/service.py
+53 -1
View File
@@ -167,6 +167,11 @@ class AgentDaemon:
lj.name = sj.get("name", lj.name)
lj.source_path = sj.get("source_path", lj.source_path)
lj.file_patterns = sj.get("file_patterns", lj.file_patterns)
lj.exclude_patterns = sj.get("exclude_patterns", lj.exclude_patterns)
lj.skip_hidden = sj.get("skip_hidden", lj.skip_hidden)
lj.skip_system = sj.get("skip_system", lj.skip_system)
lj.skip_readonly = sj.get("skip_readonly", lj.skip_readonly)
lj.sync_deletions = sj.get("sync_deletions", lj.sync_deletions)
lj.schedule_cron = sj.get("schedule_cron", lj.schedule_cron)
lj.min_stable_seconds = sj.get("min_stable_time_seconds", lj.min_stable_seconds)
@@ -191,6 +196,11 @@ class AgentDaemon:
name=sj["name"],
source_path=sj["source_path"],
file_patterns=sj["file_patterns"],
exclude_patterns=sj.get("exclude_patterns", ""),
skip_hidden=sj.get("skip_hidden", False),
skip_system=sj.get("skip_system", False),
skip_readonly=sj.get("skip_readonly", False),
sync_deletions=sj.get("sync_deletions", False),
schedule_cron=sj["schedule_cron"],
min_stable_seconds=sj["min_stable_time_seconds"],
last_status=sj.get("status", "En espera")
@@ -209,6 +219,11 @@ class AgentDaemon:
"name": lj.name,
"source_path": lj.source_path,
"file_patterns": lj.file_patterns,
"exclude_patterns": lj.exclude_patterns,
"skip_hidden": lj.skip_hidden,
"skip_system": lj.skip_system,
"skip_readonly": lj.skip_readonly,
"sync_deletions": lj.sync_deletions,
"schedule_cron": lj.schedule_cron,
"min_stable_time_seconds": lj.min_stable_seconds
}
@@ -243,11 +258,26 @@ class AgentDaemon:
logger.warning(f"Source path {source_path} for job {job_name} does not exist. Skipping.")
continue
scanner = DirectoryScanner(source_path, file_patterns, min_stable_seconds=min_stable)
scanner = DirectoryScanner(
source_path=source_path,
file_patterns=file_patterns,
exclude_patterns=getattr(job, "exclude_patterns", ""),
skip_hidden=getattr(job, "skip_hidden", False),
skip_system=getattr(job, "skip_system", False),
skip_readonly=getattr(job, "skip_readonly", False),
min_stable_seconds=min_stable
)
files = scanner.scan()
# Track files successfully backed up in this run
active_relative_paths = []
for filepath in files:
try:
rel_p = str(filepath.relative_to(Path(source_path).resolve())).replace("\\", "/")
active_relative_paths.append(rel_p)
except Exception:
pass
if not is_file_stable(filepath, min_stable_seconds=min_stable):
logger.warning(f"File {filepath.name} is currently locked or growing. Skipping.")
continue
@@ -270,6 +300,7 @@ class AgentDaemon:
res = self.uploader.upload_file(
filepath,
job_id=job.job_id,
source_path=source_path,
progress_callback=on_chunk_progress
)
logger.info(f"Successfully backed up {filepath.name}!")
@@ -284,6 +315,27 @@ class AgentDaemon:
if self.on_error:
self.on_error(filepath.name, str(ex))
# Enforce mirror replica mode (purge orphans on server)
if job.sync_deletions and job.job_id is not None:
try:
base_url = self.config.server_url.rstrip("/")
purge_payload = {
"active_relative_paths": active_relative_paths
}
resp = httpx.post(
f"{base_url}/api/jobs/{job.job_id}/purge-orphans",
headers=self._get_headers(),
json=purge_payload,
timeout=30.0
)
if resp.status_code == 200:
purge_res = resp.json()
purged_count = purge_res.get("purged_count", 0)
if purged_count > 0:
logger.info(f"Purged {purged_count} orphan files from server for job '{job_name}'")
except Exception as e:
logger.error(f"Error purging orphan files from server: {e}")
# Update job state in config after checking directory
job.last_backup_at = datetime.now(timezone.utc).strftime("%Y-%m-%d %H:%M:%S")
job.last_status = "Backup Exitoso" if job.last_status != "Error" else "Error"
windows-agent/agent/uploader.py
+10
View File
@@ -25,6 +25,7 @@ class ChunkUploader:
self,
filepath: Path,
job_id: Optional[int] = None,
source_path: Optional[str] = None,
progress_callback: Optional[Callable[[int, int, float], None]] = None,
max_retries: int = 5
) -> Dict[str, Any]:
@@ -37,6 +38,14 @@ class ChunkUploader:
full_sha256 = chunker.full_sha256
filename = filepath.name
client_relative_path = filename
if source_path:
try:
resolved_src = Path(source_path).resolve()
client_relative_path = str(filepath.relative_to(resolved_src)).replace("\\", "/")
except Exception:
pass
headers = self._get_headers()
base_url = self.config.server_url.rstrip("/")
@@ -44,6 +53,7 @@ class ChunkUploader:
# 1. Initialize or resume upload session
init_payload = {
"filename": filename,
"client_relative_path": client_relative_path,
"file_size": chunker.file_size,
"sha256": full_sha256,
"chunk_size": chunker.chunk_size,