update query

Author: StellaHuang95

analysis/kusto/01-failure-stage-breakdown.kql

@@ -1,18 +1,15 @@
-// Query 1 — Registration Failure Stage Breakdown per Manager (28 days)
-// For each manager that failed, shows WHICH stage in the registration flow broke.
-// failureStage is hierarchical: "getPipenv:nativeFinderRefresh", "constructCondaSourcingStatus", etc.
+// Query 1 — Registration Failure Stage Breakdown per Manager
+// For events that DO have failureStage, what values are present?
 // High counts at a specific stage → that code path is the priority fix target.
-let endDate = startofday(now()-1d);
 RawEventsVSCodeExt
-| where ServerTimestamp > endDate-28d and ServerTimestamp < endDate
+| where ServerTimestamp > ago(7d)
 | where EventName == "ms-python.vscode-python-envs/manager_registration.failed"
 | extend ExtVersion = tostring(Properties["common.extversion"])
-| extend _minor = toint(extract("^1\\.(\\d+)", 1, ExtVersion)), _patch = tolong(extract("^1\\.\\d+\\.(\\d+)", 1, ExtVersion))
+| extend _minor = toint(extract("^1\\.(\\d+)", 1, ExtVersion))
+| extend _patch = tolong(extract("^1\\.\\d+\\.(\\d+)", 1, ExtVersion))
 | where _minor > 23 or (_minor == 23 and _patch >= 10781012)
-| extend ManagerName = tostring(Properties.managername)
 | extend FailureStage = tostring(Properties.failurestage)
-| summarize
-    FailureCount = count(),
-    AffectedMachines = dcount(VSCodeMachineId)
-    by ManagerName, FailureStage
-| order by ManagerName asc, AffectedMachines desc
+| where isnotempty(FailureStage)
+| summarize EventCount = count(), Machines = dcount(VSCodeMachineId)
+    by ManagerName = tostring(Properties.managername), FailureStage, ExtVersion
+| order by Machines desc

analysis/kusto/02-error-type-x-failure-stage.kql

@@ -1,18 +1,19 @@
-// Query 2 — Error Type × Failure Stage Matrix (28 days)
+// Query 2 — Error Type × Failure Stage Matrix
 // Cross-tabulates errorType (what kind of error) with failureStage (where it happened).
 // This is the key diagnostic view: e.g., "connection_error at nativeFinderRefresh" means
 // PET process dies during native finder, while "tool_not_found at pathLookup" means the
 // tool binary wasn't on PATH. Prioritize cells with the highest AffectedMachines.
-let endDate = startofday(now()-1d);
 RawEventsVSCodeExt
-| where ServerTimestamp > endDate-28d and ServerTimestamp < endDate
+| where ServerTimestamp > ago(7d)
 | where EventName == "ms-python.vscode-python-envs/manager_registration.failed"
 | extend ExtVersion = tostring(Properties["common.extversion"])
-| extend _minor = toint(extract("^1\\.(\\d+)", 1, ExtVersion)), _patch = tolong(extract("^1\\.\\d+\\.(\\d+)", 1, ExtVersion))
+| extend _minor = toint(extract("^1\\.(\\d+)", 1, ExtVersion))
+| extend _patch = tolong(extract("^1\\.\\d+\\.(\\d+)", 1, ExtVersion))
 | where _minor > 23 or (_minor == 23 and _patch >= 10781012)
 | extend ManagerName = tostring(Properties.managername)
 | extend ErrorType = tostring(Properties.errortype)
 | extend FailureStage = tostring(Properties.failurestage)
+| where isnotempty(FailureStage)
 | summarize
     FailureCount = count(),
     AffectedMachines = dcount(VSCodeMachineId)

analysis/kusto/03-failure-stage-daily-trend.kql

@@ -156,9 +156,9 @@
    "metadata": {},
    "source": [
     "---\n",
-    "## 3. Daily Registration Failures by Manager & Stage (14 days)\n",
+    "## 3. Overall Setup Success Rate (28 days)\n",
     "\n",
-    "Day-by-day failure counts per manager + stage. A spike in one stage on a specific day = regression in that code path."
+    "Top-level health metric. If this drops, something is broken."
    ]
   },
   {
@@ -168,7 +168,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"03-failure-stage-daily-trend.kql\")\n",
+    "df = run_kql_file(client, \"03-overall-setup-success-rate.kql\")\n",
     "display(df)"
    ]
   },
@@ -178,41 +178,19 @@
    "metadata": {},
    "source": [
     "---\n",
-    "## 4. Overall Setup Success Rate (28 days)\n",
-    "\n",
-    "Top-level health metric. If this drops, something is broken."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "13",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "df = run_kql_file(client, \"04-overall-setup-success-rate.kql\")\n",
-    "display(df)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "14",
-   "metadata": {},
-   "source": [
-    "---\n",
-    "## 5. Manager Availability\n",
+    "## 4. Manager Availability\n",
     "\n",
     "What tools do users actually have installed? Shows registered vs skipped vs failed per manager."
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "15",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"05-manager-availability.kql\")\n",
+    "df = run_kql_file(client, \"04-manager-availability.kql\")\n",
     "display(df)\n",
     "\n",
     "if not df.empty:\n",
@@ -241,23 +219,23 @@
   },
   {
    "cell_type": "markdown",
-   "id": "16",
+   "id": "14",
    "metadata": {},
    "source": [
     "---\n",
-    "## 6. Daily Trend (14 days)\n",
+    "## 5. Daily Trend (14 days)\n",
     "\n",
     "Day-by-day trend of setup success rate. Check after shipping a new version."
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "17",
+   "id": "15",
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"06-daily-trend.kql\")\n",
+    "df = run_kql_file(client, \"05-daily-trend.kql\")\n",
     "display(df)\n",
     "\n",
     "if not df.empty:\n",
@@ -276,23 +254,23 @@
   },
   {
    "cell_type": "markdown",
-   "id": "18",
+   "id": "16",
    "metadata": {},
    "source": [
     "---\n",
-    "## 7. Error Type Distribution\n",
+    "## 6. Error Type Distribution\n",
     "\n",
     "Groups all failures by error type across setup and individual managers."
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "19",
+   "id": "17",
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"07-error-type-distribution.kql\")\n",
+    "df = run_kql_file(client, \"06-error-type-distribution.kql\")\n",
     "display(df)\n",
     "\n",
     "if not df.empty:\n",
@@ -312,52 +290,52 @@
   },
   {
    "cell_type": "markdown",
-   "id": "20",
+   "id": "18",
    "metadata": {},
    "source": [
     "---\n",
-    "## 8. Hang ↔ Failure Correlation\n",
+    "## 7. Hang ↔ Failure Correlation\n",
     "\n",
     "Do hangs always cause failures, or do some self-recover?"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "21",
+   "id": "19",
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"08-hang-failure-correlation.kql\")\n",
+    "df = run_kql_file(client, \"07-hang-failure-correlation.kql\")\n",
     "display(df)"
    ]
   },
   {
    "cell_type": "markdown",
-   "id": "22",
+   "id": "20",
    "metadata": {},
    "source": [
     "---\n",
-    "## 9. Weekly Health Summary\n",
+    "## 8. Weekly Health Summary\n",
     "\n",
     "One-stop query for weekly check. Returns all key numbers in a single row."
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "23",
+   "id": "21",
    "metadata": {},
    "outputs": [],
    "source": [
-    "df = run_kql_file(client, \"09-weekly-health-summary.kql\")\n",
+    "df = run_kql_file(client, \"08-weekly-health-summary.kql\")\n",
     "display(df)"
    ]
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": ".venv (3.13.12)",
+   "display_name": ".venv",
    "language": "python",
    "name": "python3"
   },