From 3578adb47a0de78dc5698c256a8d103857d22c33 Mon Sep 17 00:00:00 2001 From: Denys Vlasenko Date: Wed, 27 Nov 2013 13:14:24 +0100 Subject: [LIBREPORT PATCH 17/17] workflow_RHELvmcore: run analyze_VMcore too analyze_VMcore does the following: * extracts oops from vmcore-dmesg.txt (or from vmcore, if vmcore-dmesg.txt is missing); creates "kernel" element. * generates dedup hashes. * generates pkg_* elements. These steps probably aren't necessary if we take the usual step of running report_RHTSupport: RHTSupport people don't absolutely need that data. (However, I did not test whether report_RHTSupport actually works w/o analyze_VMcore). But for Machine Check Exceptions, oops extraction is crucial: without seeing "backtrace" element, which in this case shows MCE message, user won't realize he has a hardware issue, and will still contact RHTSupport, only to be told by them, after analysis, "look at your CPU/RAM, they seem to be faulty" - exactly the scenario we want to avoid (we want user to not waste his and RHTSupport's time - the problem is known). Signed-off-by: Denys Vlasenko Resolves rhbz#1059651 Signed-off-by: Jakub Filak --- src/workflows/workflow_RHELvmcore.xml.in | 1 + 1 file changed, 1 insertion(+) diff --git a/src/workflows/workflow_RHELvmcore.xml.in b/src/workflows/workflow_RHELvmcore.xml.in index 20b6989..f2a775d 100644 --- a/src/workflows/workflow_RHELvmcore.xml.in +++ b/src/workflows/workflow_RHELvmcore.xml.in @@ -5,6 +5,7 @@ collect_* + analyze_VMcore report_RHTSupport -- 1.8.3.1