# Regression in ES 7.7.0 with es-rally http\_logs Track **URL:** \ **Category:** Elasticsearch\ **Tags:** rally\ **Created:** [June 5, 2020, 1:15pm UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954 "2020-06-05T13:15:57Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![Astro-es](https://avatars.discourse-cdn.com/v4/letter/a/b782af/32.png) [@Astro-es](https://discuss.elastic.co/u/Astro-es)\ **Post date:** [June 5, 2020, 1:15pm UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/1 "2020-06-05T13:15:58Z")
Elasticsearch 7.7.0 shows regression when benchmarked against 7.2.0 for the following challenges in es-rally: Es-Rally Version: 2.0 Track: http\_logs Challenge: append-sorted-no-conflicts Total Regressions: 18 ES Versions: 7.7.0 vs 7.2.0 Total ES nodes: 2 (Data, Master) ES Instance Type: 16 cores, 64GB The regression was consistently observed for this specific challenge in http\_logs track. The 7.7.0 shows improvements in the other challenges like: append-no-conflicts and append-no-conflicts-index-only Please check. ---
**Author:** ![Christian\_Dahlqvist](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/christian_dahlqvist/32/4617_2.png) [@Christian\_Dahlqvist](https://discuss.elastic.co/u/Christian_Dahlqvist)\ **Post date:** [June 23, 2020, 4:58am UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/2 "2020-06-23T04:58:48Z")
I had a look at the [results of the nightly regression tests](https://elasticsearch-benchmark-analytics.elastic.co/app/kibana#/dashboard/d02cc8e5-3f42-482a-a3d9-d09956db3040?_g=()&_a=(description:'',filters:!(),fullScreenMode:!f,options:(darkTheme:!f),panels:!((gridData:(h:32,i:'1',w:48,x:0,y:0),id:'73bd349a-2198-4590-94ee-1cab5fec2c3e',panelIndex:'1',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'2',w:24,x:0,y:32),id:'8c877fe1-8dce-4fa9-af8c-bec5b07057ac',panelIndex:'2',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'3',w:24,x:24,y:32),id:'43948549-1bcf-4952-aa71-1d4636a9900b',panelIndex:'3',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'4',w:24,x:0,y:64),id:'87765038-bcb4-45c3-8598-137056f00a81',panelIndex:'4',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'5',w:24,x:24,y:64),id:'43a4a7b5-e86c-4ede-9772-8e0b5006aa94',panelIndex:'5',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'6',w:24,x:0,y:96),id:'25535079-f745-4c0d-b571-189c78bea4c1',panelIndex:'6',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'7',w:24,x:24,y:96),id:'532eca54-5d57-404e-ad6c-5429ce6168be',panelIndex:'7',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'8',w:24,x:0,y:128),id:'91e94c07-480d-4162-8094-64adbb1f526a',panelIndex:'8',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'9',w:24,x:24,y:128),id:'1ec9dcfd-85c6-4121-8677-542ea4b3aee8',panelIndex:'9',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'10',w:24,x:0,y:160),id:'1180fff6-da75-48cf-ae96-00a92e018c52',panelIndex:'10',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'11',w:24,x:24,y:160),id:e8872fe8-eefc-4132-be8b-83e85a7740cb,panelIndex:'11',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'12',w:24,x:0,y:192),id:'3fbc1e41-7870-4c06-91cc-bc84cc4c446c',panelIndex:'12',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'13',w:24,x:24,y:192),id:'9a47a754-c6d3-4284-a897-4addcdc4dea7',panelIndex:'13',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'14',w:24,x:0,y:224),id:'92bdeb81-6025-4266-bce8-af742d801885',panelIndex:'14',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'15',w:24,x:24,y:224),id:'277e9d7f-d51f-499f-ab6b-f51b05e01c75',panelIndex:'15',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'16',w:24,x:0,y:256),id:ad392600-e69e-4be8-af79-a9310c94d238,panelIndex:'16',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'17',w:24,x:24,y:256),id:efab7c77-feb2-4018-886f-c43e6e8cb9b9,panelIndex:'17',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'18',w:24,x:0,y:288),id:c37dc20d-c53f-44f1-ad52-df1c2770d340,panelIndex:'18',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'19',w:24,x:24,y:288),id:'5cef9d83-12ff-4c34-88d2-1bc717b877f5',panelIndex:'19',type:visualization,version:'6.3.2'),(gridData:(h:32,i:'20',w:24,x:0,y:320),id:d9c8b795-cab0-4766-a816-2d615201126f,panelIndex:'20',type:visualization,version:'6.3.2')),query:(language:lucene,query:(query_string:(analyze_wildcard:!t,query:'*'))),timeRestore:!f,title:nightly-default-http_logs,viewMode:view)) and did not see any regression. I do however not have the details exactly what is behind these numbers. Can you perhaps provide some additional information about the environment where you saw this regression? ---
**Author:** ![Astro-es](https://avatars.discourse-cdn.com/v4/letter/a/b782af/32.png) [@Astro-es](https://discuss.elastic.co/u/Astro-es)\ **Post date:** [June 30, 2020, 5:39am UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/3 "2020-06-30T05:39:48Z")
The regression was observed repeatedly in our rally races with track http\_logs track and challenge append-sorted-no-conflicts. The environment details are provided in the question itself. Thanks! ---
**Author:** ![Christian\_Dahlqvist](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/christian_dahlqvist/32/4617_2.png) [@Christian\_Dahlqvist](https://discuss.elastic.co/u/Christian_Dahlqvist)\ **Post date:** [June 30, 2020, 6:01am UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/4 "2020-06-30T06:01:41Z")
Are you using local SSDs? Indexing is I/O intensive so it would be good to get a full picture of how your environment compares to the one running the nightly benchmarks. Is this the only benchmark where you see a regression? Are all other benchmarks consistent with previous versions in your environment? ---
**Author:** ![Astro-es](https://avatars.discourse-cdn.com/v4/letter/a/b782af/32.png) [@Astro-es](https://discuss.elastic.co/u/Astro-es)\ **Post date:** [June 30, 2020, 6:44am UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/5 "2020-06-30T06:44:59Z")
> [@Christian\_Dahlqvist](#): > > Is this the only benchmark where you see a regression? Are all other benchmarks consistent with previous versions in your environment? - Yes, all races see performance improvements with 7.7.0 except the http\_logs track with challenge append-sorted-no-conflicts > Are you using local SSDs? Indexing is I/O intensive so it would be good to get a full picture of how your environment compares to the one running the nightly benchmarks - Not using local SSD's. Performance is good enough for our requirements ---
**Author:** ![Christian\_Dahlqvist](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/christian_dahlqvist/32/4617_2.png) [@Christian\_Dahlqvist](https://discuss.elastic.co/u/Christian_Dahlqvist)\ **Post date:** [June 30, 2020, 6:48am UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/6 "2020-06-30T06:48:25Z")
Can you monitor disk utilization and iowait for this benchmark and compare the two versions? I wonder if this benchmark might have resulted in increased I/O, which has a more notable effect in your cluster than the one that runs nightly benchmarks as it is using fast, local SSDs. @dliappis Are you monitoring and comparing resource usage, e.g. disk IOPS, for the nightly benchmarks? ---
**Author:** ![dliappis](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/dliappis/32/56174_2.png) [@dliappis](https://discuss.elastic.co/u/dliappis)\ **Post date:** [July 1, 2020, 2:23pm UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/7 "2020-07-01T14:23:47Z")
> [@Christian\_Dahlqvist](#): > > @dliappis Are you monitoring and comparing resource usage, e.g. disk IOPS, for the nightly benchmarks? Not in a systematic way yet. Reading the thread, one major change in Elasticsearch 7.7.0 is the change to Java 14 which forced the use of G1GC (by removing the CMS garbage collector); do you use the bundled JDK or explicitly setting your own `JAVA_HOME`? I suspect the former, but anyway looking at the release benchmarks for `7.8.0` in [Elasticsearch Benchmarks](https://elasticsearch-benchmarks.elastic.co/#tracks/http-logs/release) (see `sorted-4g` median indexing throughput) I don't see a significant regression even if the heap is a bit reserved at `4g`. ---
**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\ **Post date:** [July 29, 2020, 2:23pm UTC](https://discuss.elastic.co/t/regression-in-es-7-7-0-with-es-rally-http-logs-track/235954/8 "2020-07-29T14:23:48Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.