mirrors/pikasTech-unidesk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: report remote gc target shortfall

Browse Source
This commit is contained in:
Codex
2026-06-02 00:26:28 +00:00
parent bc679b1547
commit 8be605aac3
4 changed files with 110 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scripts/src/gc-remote.ts
+95 -15
View File
@@ -20,6 +20,7 @@ interface RemoteGcOptions {
registryGcOnly: boolean;
registryKeepPerRepo: number;
registryMinAgeHours: number;
targetUsePercent?: number;
jobId?: string;
limit: number;
resultLimit: number;
@@ -109,6 +110,10 @@ function parseRemoteGcOptions(args: string[]): RemoteGcOptions {
} else if (arg === "--registry-min-age-hours") {
const value = parseNonNegativeNumber(arg, args[++index]);
options.registryMinAgeHours = value;
} else if (arg === "--target-use-percent") {
const value = parseNonNegativeNumber(arg, args[++index]);
if (!Number.isInteger(value) || value < 1 || value > 99) throw new Error("--target-use-percent must be an integer from 1 to 99");
options.targetUsePercent = value;
} else if (arg === "--job-id") {
const value = args[++index];
if (!value || !/^[A-Za-z0-9._-]{1,128}$/u.test(value)) throw new Error("--job-id must be a safe job id");
@@ -529,6 +534,16 @@ def fmt_bytes(value):
idx += 1
return ("%0.0f %s" if size >= 10 or idx == 0 else "%0.1f %s") % (size, units[idx])
def disk_use_percent(size_bytes, used_bytes):
try:
size = int(size_bytes or 0)
used = int(used_bytes or 0)
except Exception:
return None
if size <= 0:
return None
return int((max(0, used) * 100 + size - 1) // size)
def parse_journal_usage(text):
m = re.search(r"take up\s+([0-9.]+)\s*([KMGT]?)(?:i?B|B)?", text, re.I)
if not m:
@@ -1485,7 +1500,66 @@ def collect_candidates(observed_at):
})
return sorted(candidates, key=lambda item: item.get("estimatedReclaimBytes") or 0, reverse=True)
def summarize(candidates, returned):
def target_assessment(disk, estimated_reclaim):
raw = OPTIONS.get("targetUsePercent")
if raw is None:
return None
if not disk:
return {
"targetUsePercent": raw,
"ok": False,
"state": "unavailable",
"reason": "disk-snapshot-unavailable",
}
try:
target = int(raw)
size = int(disk.get("sizeBytes") or 0)
used = int(disk.get("usedBytes") or 0)
reclaim = max(0, int(estimated_reclaim or 0))
except Exception:
return {
"targetUsePercent": raw,
"ok": False,
"state": "unavailable",
"reason": "invalid-disk-snapshot",
}
target_used_bytes = (size * target) // 100
required = max(0, used - target_used_bytes)
projected_used = max(0, used - reclaim)
projected_use_percent = disk_use_percent(size, projected_used)
enough = reclaim >= required
if required == 0:
state = "already-below-target"
elif enough:
state = "candidate-estimate-meets-target"
elif reclaim <= 0:
state = "safe-stop-no-meaningful-candidates"
else:
state = "shortfall"
shortfall = max(0, required - reclaim)
return {
"targetUsePercent": target,
"ok": required == 0 or enough,
"state": state,
"currentUsePercent": disk.get("usePercent"),
"currentUsedBytes": used,
"currentUsed": fmt_bytes(used),
"targetUsedBytes": target_used_bytes,
"targetUsed": fmt_bytes(target_used_bytes),
"requiredReclaimBytes": required,
"requiredReclaim": fmt_bytes(required),
"estimatedReclaimBytes": reclaim,
"estimatedReclaim": fmt_bytes(reclaim),
"shortfallBytes": shortfall,
"shortfall": fmt_bytes(shortfall),
"projectedUsedBytes": projected_used,
"projectedUsed": fmt_bytes(projected_used),
"projectedUsePercent": projected_use_percent,
"safeStop": required > 0 and not enough,
"decision": "stop-and-escalate-retention-or-capacity" if required > 0 and not enough else "target-covered-by-safe-candidates",
}
def summarize(candidates, returned, disk=None):
by_kind = {}
total = 0
for item in candidates:
@@ -1505,6 +1579,7 @@ def summarize(candidates, returned):
"returnedEstimatedReclaimBytes": returned_total,
"returnedEstimatedReclaim": fmt_bytes(returned_total),
"byKind": by_kind,
"target": target_assessment(disk, total),
}
def assert_tmp_candidate(path):
@@ -1604,6 +1679,7 @@ def returned_results(results):
return (failed + started + succeeded)[:int(OPTIONS.get("resultLimit") or 50)]
def plan_payload(observed_at, preflight, protected, candidates, visible):
disk = df_snapshot()
return {
"ok": True,
"action": "gc remote plan",
@@ -1612,9 +1688,9 @@ def plan_payload(observed_at, preflight, protected, candidates, visible):
"mutation": False,
"observedAt": observed_at,
"options": OPTIONS,
"diskBefore": df_snapshot(),
"diskBefore": disk,
"clusterPreflight": preflight,
"summary": summarize(candidates, visible),
"summary": summarize(candidates, visible, disk),
"candidates": visible,
"protected": protected,
"policy": {
@@ -1636,6 +1712,7 @@ def plan_payload(observed_at, preflight, protected, candidates, visible):
"HWLAB DEV runtime and local-path PVC data are protected and require HWLAB-specific retention commands.",
"Core dump cleanup only removes untracked /root/unidesk/core.<pid> regular files with no active fuser reference.",
"HWLAB registry retention is opt-in: it keeps workload tag/digest refs, all tags newer than the retention age and the newest N tags per repo before official registry garbage-collect.",
"When summary.target.safeStop is true, do not broaden deletion scope; choose registry retention, k3s/containerd image cache maintenance, PVC/runtime retention or capacity expansion explicitly.",
],
},
}
@@ -1684,6 +1761,20 @@ def main():
disk_after = df_snapshot()
failed = [item for item in results if item.get("status") == "failed"]
returned = returned_results(results)
run_summary = summarize(visible, returned, disk_before)
run_summary.update({
"plannedCandidateCount": len(visible),
"attemptedCount": len(results),
"startedCount": len([item for item in results if item.get("status") == "started"]),
"succeededCount": len([item for item in results if item.get("status") == "succeeded"]),
"failedCount": len(failed),
"actualDiskReclaimBytes": (disk_after["availableBytes"] - disk_before["availableBytes"]) if disk_before and disk_after else None,
"actualDiskReclaim": fmt_bytes(disk_after["availableBytes"] - disk_before["availableBytes"]) if disk_before and disk_after else None,
"targetAfter": target_assessment(disk_after, 0),
"resultCount": len(results),
"returnedResultCount": len(returned),
"omittedResultCount": max(0, len(results) - len(returned)),
})
payload = {
"ok": len(failed) == 0,
"action": "gc remote run",
@@ -1696,18 +1787,7 @@ def main():
"diskAfter": disk_after,
"clusterPreflight": preflight,
"clusterAfter": cluster_preflight(),
"summary": {
"plannedCandidateCount": len(visible),
"attemptedCount": len(results),
"startedCount": len([item for item in results if item.get("status") == "started"]),
"succeededCount": len([item for item in results if item.get("status") == "succeeded"]),
"failedCount": len(failed),
"estimatedReclaimBytes": sum(int(item.get("estimatedReclaimBytes") or 0) for item in visible),
"actualDiskReclaimBytes": (disk_after["availableBytes"] - disk_before["availableBytes"]) if disk_before and disk_after else None,
"resultCount": len(results),
"returnedResultCount": len(returned),
"omittedResultCount": max(0, len(results) - len(returned)),
},
"summary": run_summary,
"results": returned,
"protected": protected,
}
scripts/src/help.ts
+2
View File
@@ -280,6 +280,7 @@ function gcHelp(): unknown {
"bun scripts/cli.ts gc policy plan",
"bun scripts/cli.ts gc policy install",
"bun scripts/cli.ts gc remote G14 plan",
"bun scripts/cli.ts gc remote G14 plan --target-use-percent 50 --include-hwlab-registry",
"bun scripts/cli.ts gc remote G14 run --confirm",
"bun scripts/cli.ts gc remote G14 status --job-id <id>",
"bun scripts/cli.ts gc plan --full",
@@ -306,6 +307,7 @@ function gcHelp(): unknown {
"--registry-gc-only": "remote G14 only: run official registry garbage-collect without deleting additional tags; intended for interrupted registry retention recovery",
"--registry-keep-per-repo N": "remote registry only: keep at least N newest tags per service repo; default 20, minimum 1",
"--registry-min-age-hours N": "remote registry only: keep all tags newer than N hours; default 48, minimum 0",
"--target-use-percent N": "remote only: evaluate whether planned candidates can reduce root filesystem use to N%; reports required reclaim, projected use, shortfall and safe-stop decision",
"--job-id ID": "remote status only: inspect a long-running remote gc job",
"--limit N": "number of candidates returned and executed by run when --full is not set; default 50",
"--result-limit N": "number of per-candidate run results returned when --full is not set; default 50",