Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[improvement](inverted index) Disable the use of skipping write index on load #34719

Merged
merged 1 commit into from May 13, 2024

Conversation

qidaye
Copy link
Contributor

@qidaye qidaye commented May 11, 2024

Proposed changes

When skip_write_index_on_load is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use inverted_index_ram_dir_enable = true and inverted_index_storage_format=V2 to reduce IO and CPU consumption. So we disable it now.

  1. Disable setting skip_write_index_on_load to true in create table stmt.
  2. Disable setting skip_write_index_on_load to true in alter table properties stmt. You can still alter skip_write_index_on_load to false.

Further comments

If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...

… on load

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.
@doris-robot
Copy link

Thank you for your contribution to Apache Doris.
Don't know what should be done next? See How to process your PR

Since 2024-03-18, the Document has been moved to doris-website.
See Doris Document.

@qidaye
Copy link
Contributor Author

qidaye commented May 11, 2024

run buildall

@doris-robot
Copy link

TPC-H: Total hot run time: 40555 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 687169bffc483872b117d6b990bf5ff210c5bde3, data reload: false

------ Round 1 ----------------------------------
q1	17610	4391	4291	4291
q2	2019	192	198	192
q3	11119	1256	1184	1184
q4	10953	789	720	720
q5	7541	2677	2671	2671
q6	221	136	135	135
q7	1000	597	574	574
q8	9575	2111	2063	2063
q9	9227	7076	6583	6583
q10	8951	3691	3698	3691
q11	458	236	259	236
q12	493	216	220	216
q13	17776	2959	2927	2927
q14	264	225	222	222
q15	508	463	462	462
q16	503	383	372	372
q17	950	672	677	672
q18	7969	7474	7363	7363
q19	6019	1542	1514	1514
q20	647	311	318	311
q21	4968	3878	3907	3878
q22	363	281	278	278
Total cold run time: 119134 ms
Total hot run time: 40555 ms

----- Round 2, with runtime_filter_mode=off -----
q1	4354	4227	4209	4209
q2	370	265	264	264
q3	2944	2704	2721	2704
q4	1810	1553	1550	1550
q5	5227	5252	5249	5249
q6	211	122	122	122
q7	2203	1906	1872	1872
q8	3177	3294	3405	3294
q9	8325	8353	8335	8335
q10	3847	3672	3703	3672
q11	572	470	489	470
q12	748	600	591	591
q13	17317	2932	2986	2932
q14	277	273	260	260
q15	512	465	471	465
q16	472	399	410	399
q17	1759	1467	1470	1467
q18	7505	7353	7492	7353
q19	1665	1542	1532	1532
q20	1972	1780	1751	1751
q21	4822	4682	4898	4682
q22	560	539	504	504
Total cold run time: 70649 ms
Total hot run time: 53677 ms

@doris-robot
Copy link

TPC-DS: Total hot run time: 187611 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpcds-tools
TPC-DS sf100 test result on commit 687169bffc483872b117d6b990bf5ff210c5bde3, data reload: false

query1	906	365	360	360
query2	7193	2350	2439	2350
query3	6663	214	220	214
query4	22756	21267	21264	21264
query5	4163	429	416	416
query6	262	175	175	175
query7	4595	292	284	284
query8	242	206	189	189
query9	8528	2413	2388	2388
query10	435	245	266	245
query11	14623	14266	14079	14079
query12	140	91	87	87
query13	1650	372	369	369
query14	10428	8424	8319	8319
query15	251	170	177	170
query16	8179	282	263	263
query17	1842	560	540	540
query18	2090	274	263	263
query19	262	148	142	142
query20	90	81	86	81
query21	191	132	128	128
query22	5108	4963	4961	4961
query23	34297	33518	33446	33446
query24	12022	2874	2835	2835
query25	651	381	357	357
query26	1743	152	153	152
query27	3011	316	319	316
query28	7645	2045	2038	2038
query29	1005	604	594	594
query30	291	147	154	147
query31	967	737	753	737
query32	91	52	53	52
query33	750	248	240	240
query34	1097	484	474	474
query35	812	671	656	656
query36	1061	907	936	907
query37	287	66	66	66
query38	2882	2795	2756	2756
query39	1651	1683	1565	1565
query40	274	124	124	124
query41	43	38	38	38
query42	128	94	93	93
query43	583	559	537	537
query44	1225	707	724	707
query45	266	243	251	243
query46	1071	746	707	707
query47	1994	1904	1928	1904
query48	377	286	292	286
query49	1189	388	394	388
query50	765	385	388	385
query51	6799	6755	6778	6755
query52	103	87	94	87
query53	347	275	276	275
query54	1008	429	424	424
query55	75	78	72	72
query56	239	227	216	216
query57	1267	1143	1146	1143
query58	217	198	200	198
query59	3613	3153	3255	3153
query60	285	232	237	232
query61	95	89	90	89
query62	681	474	467	467
query63	309	282	278	278
query64	9762	7368	7326	7326
query65	3167	3108	3071	3071
query66	1393	366	345	345
query67	15433	15253	14918	14918
query68	4821	538	529	529
query69	532	304	304	304
query70	1154	1161	1129	1129
query71	447	323	262	262
query72	7951	2560	2425	2425
query73	716	318	325	318
query74	6484	6125	6109	6109
query75	3693	2642	2662	2642
query76	3604	943	932	932
query77	632	267	261	261
query78	10547	10197	9992	9992
query79	2513	511	510	510
query80	945	436	436	436
query81	524	216	219	216
query82	1553	94	94	94
query83	265	168	157	157
query84	230	85	80	80
query85	1574	274	314	274
query86	477	317	314	314
query87	3345	3134	3124	3124
query88	4125	2392	2398	2392
query89	463	392	377	377
query90	1977	180	183	180
query91	123	98	99	98
query92	59	47	49	47
query93	2086	507	484	484
query94	1221	174	180	174
query95	390	297	306	297
query96	609	270	264	264
query97	3147	2999	3016	2999
query98	228	220	213	213
query99	1153	896	892	892
Total cold run time: 291156 ms
Total hot run time: 187611 ms

Copy link
Contributor

@xiaokang xiaokang left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@github-actions github-actions bot added the approved Indicates a PR has been approved by one committer. label May 11, 2024
Copy link
Contributor

PR approved by at least one committer and no changes requested.

Copy link
Contributor

PR approved by anyone and no changes requested.

@qidaye
Copy link
Contributor Author

qidaye commented May 13, 2024

run p0

1 similar comment
@qidaye
Copy link
Contributor Author

qidaye commented May 13, 2024

run p0

Copy link
Member

@eldenmoon eldenmoon left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@qidaye qidaye merged commit ef05e13 into apache:master May 13, 2024
27 of 30 checks passed
@qidaye qidaye deleted the disable_skip_write_index_on_load branch May 13, 2024 07:27
nextdreamblue pushed a commit to nextdreamblue/doris that referenced this pull request May 13, 2024
… on load (apache#34719)

When `skip_write_index_on_load` is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use `inverted_index_ram_dir_enable = true` and `inverted_index_storage_format=V2` to reduce IO and CPU consumption. So we disable it now.

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.

Co-authored-by: Luennng <luennng@gmail.com>
M1saka2003 pushed a commit to M1saka2003/doris that referenced this pull request May 14, 2024
… on load (apache#34719)

When `skip_write_index_on_load` is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use `inverted_index_ram_dir_enable = true` and `inverted_index_storage_format=V2` to reduce IO and CPU consumption. So we disable it now.

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.

Co-authored-by: Luennng <luennng@gmail.com>
ByteYue pushed a commit to ByteYue/doris that referenced this pull request May 15, 2024
… on load (apache#34719)

When `skip_write_index_on_load` is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use `inverted_index_ram_dir_enable = true` and `inverted_index_storage_format=V2` to reduce IO and CPU consumption. So we disable it now.

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.

Co-authored-by: Luennng <luennng@gmail.com>
yiguolei pushed a commit that referenced this pull request May 18, 2024
… on load (#34719)

When `skip_write_index_on_load` is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use `inverted_index_ram_dir_enable = true` and `inverted_index_storage_format=V2` to reduce IO and CPU consumption. So we disable it now.

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.

Co-authored-by: Luennng <luennng@gmail.com>
M1saka2003 pushed a commit to M1saka2003/doris that referenced this pull request May 24, 2024
… on load (apache#34719)

When `skip_write_index_on_load` is turned on, users will get an error when querying for the latest data(not compacted), giving them a bad experience. And we can use `inverted_index_ram_dir_enable = true` and `inverted_index_storage_format=V2` to reduce IO and CPU consumption. So we disable it now.

1. Disable setting `skip_write_index_on_load` to `true` in create table stmt.
2. Disable setting `skip_write_index_on_load` to `true` in alter table properties stmt. You can still alter `skip_write_index_on_load` to `false`.

Co-authored-by: Luennng <luennng@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
approved Indicates a PR has been approved by one committer. reviewed
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

5 participants