Log Analysis Agent

log analysis · ready

Every failure is a lesson your logs already wrote

Choose a log file to scan failures, exceptions, slow steps and root causes — explained, not just listed.

failed steps

—

Failures traced across the loaded log.

slow steps

—

Steps running slower than their baseline.

root cause

—

High-confidence root cause from the failures and slowdowns.

Incident snapshot

Load a log to see today's incidents

Lines where an Action/Step logs a failing Return Code (any negative RC, or 400/401/403/404/500/503/504) or a Selected OUTCOME of FAILURE / ERROR / UNSPECIFIED_FAILURE. Click for the full list.?0 active

Failed steps

No log loaded

Steps whose logged "completed in N msec" is ≥ 3s, plus VTAG DataTransform views ≥ 3s. The detail view bands them ≥3s / ≥5s / ≥9s.?0 slow

Slow steps

No log loaded

For each distinct failure, the analyzer scans UP from the failure line (within the same MC_ID) to find the originating error — ORA-/SQL, XSLT, exception, timeout, validation or dependency — and states the probable cause with evidence.?analysis

Root cause

Run analysis

Lines containing an exception marker — Exception, Caused by:, NullPointerException, SocketTimeoutException or I/O error. Click for the full list grouped by action → step.?0 found

Exceptions

No log loaded

The failure lines (same source as the Failed steps card), grouped by Action → Step to show where failures concentrate — with count, return code and last-seen time per group.?0 steps

Failure hotspots

No log loaded

Count of unique authoritative correlation IDs (MC_ID, field 8 of the structured line). Bare 7–12 digit numbers inside XML payloads are excluded, so the count isn't inflated.?0 ids

Affected MCIDs

No log loaded

All parsed log lines grouped by MC_ID (or Username) — line count, error count and last-seen time per group. Click to explore that group in Discover.?groups

Activity by MC_ID

No log loaded

Notable problems surfaced from the scan — derived from the failure, exception and slow-step counts found in the loaded log, with severity and the MCID/time they occurred.?0 active

Alerts

No log loaded

Suggested config/action fixes derived from the probable-cause categories in Root Cause and the Action → Step where each failure originates.?AI

Recommendations

Ask AI for fixes

Every line in the loaded log (the server's line count). For a multi-file selection it's the combined total of all selected files. Opens the full log viewer.?0 lines

Log volume

No log loaded

Per-occurrence VTAG DataTransform view times ALIS logs directly ("Total Time Taken to Process View(...)"). The number is how many transform occurrences were timed; click for the full report (per VTAG / interface / MCID + slowest occurrences).?0 transforms

VTAG timings

No log loaded

Local utilities that run without a loaded log — XSLT transform (with ALIS Java stub mode), XML compare, SQL compare, XML/XSLT beautify and SQL format.?5 tools

Tools

XSLT · XML/SQL compare · XML/SQL format

Use 📁 Choose log at the top-right to select a log file.

Every line in the loaded log (the server's line count). For a multi-file selection it's the combined total of all selected files.?

Total Lines

—

Lines where an Action/Step logs a failing Return Code (any negative RC, or 400/401/403/404/500/503/504) or a Selected OUTCOME of FAILURE / ERROR / UNSPECIFIED_FAILURE. Click the card for the full list.?

Failures 🔎

—

Lines containing an exception marker — Exception, Caused by:, NullPointerException, SocketTimeoutException or I/O error. Click for the full list grouped by action → step.?

Exceptions 🔎

—

Steps whose logged "completed in N msec" is ≥ 3s, plus VTAG DataTransform views ≥ 3s. The detail view bands them ≥3s / ≥5s / ≥9s.?

Slow Steps 🔎

—

Per-occurrence VTAG DataTransform view times that ALIS logs directly ("Total Time Taken to Process View(...)"). The number is how many transform occurrences were timed; click for the full report (per VTAG / interface / MCID, with the slowest occurrences).?

VTAG Timings 🔎

—

Count of unique authoritative correlation IDs (MC_ID, field 8 of the structured line). Bare 7–12 digit numbers inside XML payloads are excluded, so the count isn't inflated.?

MCIDs 🔎

—

Failure Hotspots

Where failures concentrate

The failure lines (same source as the Failures card), grouped by Action → Step to show where failures concentrate — with count, return code and last-seen time per group.? —

#	Action / Step	RC	Count	Last
Load a log to see failing steps

Activity by MC_ID

Top groups · click to explore

All parsed log lines grouped by MC_ID (or Username) — line count, error count and last-seen time per group. Toggle MC_ID / Username; click a row to explore that group in Discover.?

MC_ID	Lines	Err	Last
Load a log to see groups

Recent log entries

A live preview of the most recent parsed log lines (the de-duplicated tail of the log). The LEVEL (INFO / WARN / ERROR) is inferred from keywords in each line; click any row to drill into that line in Discover.?

Live

Timestamp	Service	Level	Message	MCID	Line
Loading...

—

Alerts

0 active

Notable problems surfaced from the scan — derived from the failure, exception and slow-step counts found in the loaded log, with severity and the MCID/time they occurred.?

SeverityDescriptionTimeMCIDStatus

✓ No alerts

Root cause

Analysis pending

For each distinct failure, the analyzer scans UP from the failure line (staying within the same MC_ID) to find the originating error — ORA-/SQL, XSLT, exception, timeout, validation or dependency — and states the probable cause with the evidence line.?

Run analysis to see root causes

Recommendations

Suggested fixes

Suggested config/action fixes derived from the probable-cause categories in Root Cause and the Action → Step where each failure originates. Use "AI fix" for tailored suggestions.?

Ask AI for recommendations

▸	TIMESTAMP	CATEGORY	LEVEL	USERNAME	MC_ID	CT_ID	RESP (MS)	MESSAGE
Select a log file to explore.

Input XML

XSLT

Stub ALIS Java functions (JNDI, etc.)

Output

Pretty-print

Output will appear here after you Submit.