From 1d275959cb46cadaabd19635fe29fc154c9bf7f8 Mon Sep 17 00:00:00 2001 From: zhangdebo Date: Wed, 23 May 2018 17:02:23 +0800 Subject: [PATCH] Fix typo Change-Id: Idc204b736ec2352c96b1303f2aaf1f222ff5d1f4 --- doc/source/usage.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/source/usage.rst b/doc/source/usage.rst index 0f7b7be36..ddb7a9369 100644 --- a/doc/source/usage.rst +++ b/doc/source/usage.rst @@ -190,7 +190,7 @@ percentile for success across at least two stes of hardware. Adjust as required. Note that `batch_size` can not produce a batch of unintrospected ndoes if none exist so the last batch may be below the maximum size. When nodes in a batch fail the `failure_count` is incremented and the nodes are returned to the pool. So it's possible that same node will -fail again in another batch. There is a saftey mechanism that will kill Yoda if a node exceeds +fail again in another batch. There is a safety mechanism that will kill Yoda if a node exceeds 10 retries as that's pretty much garunteed to be misconfigured. For bulk introspection all nodes are tried once and what you get is what you get.