From 56aba17efee5b3ad4c6fdaf5b95779cc27e29aea Mon Sep 17 00:00:00 2001 From: Pius Date: Tue, 5 Oct 2021 17:53:11 -0700 Subject: [PATCH] Add aggregations/CCS memory leak as a 7.14.x known issue There is currently no plans to back port https://github.com/elastic/elasticsearch/pull/78404 to 7.14.x. Adding this as a known issue to the release notes for 7.14.x per discussion with @qhoxie . Please also back port this doc change to 7.14.0 and 7.14.1. Thx! --- docs/reference/release-notes/7.14.asciidoc | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/docs/reference/release-notes/7.14.asciidoc b/docs/reference/release-notes/7.14.asciidoc index 5e8f3d4a38678..da025769e256d 100644 --- a/docs/reference/release-notes/7.14.asciidoc +++ b/docs/reference/release-notes/7.14.asciidoc @@ -3,6 +3,22 @@ Also see <>. +[[known-issues-7.14.2]] +[discrete] +=== Known issues + +* Aggregations: When a cross cluster search request is proxied, the memory for the aggregations on the +proxy node will not be freed. The trigger is cross cluster search using aggregations where minimize +roundtrips is not effective (e.g. when minimize roundtrips is explicitly disabled, or implicitly disabled +when using scroll, async and point in time searches). This issue can happen in all modes of remote connections +configured for cross cluster search (sniff and proxy). In sniff mode, we only connect to a subset of the +remote nodes (by default 3). So if the remote node we want to send a request to is not one of those 3, +we must send the request as a proxy request. The workaround is to periodically restart nodes with heap pressure +or cancel the aggregation tasks on the nodes. ++ +We have fixed this issue in {es} 7.15.1 and later versions. For more details, +see {es-pull}78404[#78404]. + [[enhancement-7.14.2]] [float] === Enhancements