From 2fec66dfdcf8bc7c901dc399276ae14f825610f4 Mon Sep 17 00:00:00 2001 From: i-robot Date: Thu, 30 Sep 2021 10:54:57 +0000 Subject: [PATCH] !1155 English language fix in spill.md * English language fix in spill.md --- hetu-docs/en/admin/spill.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/hetu-docs/en/admin/spill.md b/hetu-docs/en/admin/spill.md index 6f578019b..8232e2a56 100644 --- a/hetu-docs/en/admin/spill.md +++ b/hetu-docs/en/admin/spill.md @@ -12,7 +12,7 @@ Properties related to spilling are described in `tuning-spilling`. ## Memory Management and Spill -By default, openLooKeng kills queries if the memory requested by the query execution exceeds session properties `query_max_memory` or `query_max_memory_per_node`. This mechanism ensures fairness in allocation of memory to queries and prevents deadlock caused by memory allocation. It is efficient when there is a lot of small queries in the cluster, but leads to killing large queries that don\'t stay within the limits. +By default, openLooKeng kills queries if the memory requested by the query execution exceeds session properties `query_max_memory` or `query_max_memory_per_node`. This mechanism ensures fairness in allocation of memory to queries and prevents deadlock caused by memory allocation. It is efficient when there are lots of small queries in the cluster, but leads to killing large queries that don\'t stay within the limits. To overcome this inefficiency, the concept of revocable memory was introduced. A query can request memory that does not count toward the limits, but this memory can be revoked by the memory manager at any time. When memory is revoked, the query runner spills intermediate data from memory to disk and continues to process it later.