From 993f0cb0c6107e571acfbe45d03ef46ffbe473ab Mon Sep 17 00:00:00 2001
From: "Andrea Frittoli (andreaf)" <andrea.frittoli@hp.com>
Date: Tue, 19 May 2015 12:19:20 -0700
Subject: [PATCH] Add query for bug 1456771

Since May 15th 2015 I sometimes see failures in both
check and gate pipelines with "Failure prepping block
device in the gate".

2771 hits in 7 days, check and gate, not 100% failure
because this is a delete path which is not explicitly
tested for most scenarios, so it's only really a "fail"
in explicit delete server tests.

https://review.openstack.org/#/c/145738/ appears to
have introduced the race.

Related-Bug: #1456771

Change-Id: I1e9392067f19f1fbd86c5407debf936a9fdf8228
---
 queries/1456771.yaml | 4 ++++
 1 file changed, 4 insertions(+)
 create mode 100644 queries/1456771.yaml

diff --git a/queries/1456771.yaml b/queries/1456771.yaml
new file mode 100644
index 00000000..c26a98e3
--- /dev/null
+++ b/queries/1456771.yaml
@@ -0,0 +1,4 @@
+query: >-
+  message:"_default_block_device_names" AND
+  message:"TypeError: 'NoneType' object has no attribute '__getitem__'" AND
+  tags:"screen-n-cpu.txt"