Skip to content

Conversation

@sollhui
Copy link
Contributor

@sollhui sollhui commented Mar 15, 2024

Proposed changes

There is too param control routine load task:
FE config: max_routine_load_task_num_per_be
BE config: routine_load_thread_pool_size

  1. Users have to modify both when using routine load.
  2. The parameters of be do not support dynamic modification. When the user's routing load needs to be modified too much, they have to restart be

Based on these two considerations, the logical control of the number of tasks on the BE end was optimized, and the parameters were removed routine_load_thread_pool_size.

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...

@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

@sollhui
Copy link
Contributor Author

sollhui commented Mar 15, 2024

run buildall

@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

1 similar comment
@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@sollhui
Copy link
Contributor Author

sollhui commented Mar 15, 2024

run buildall

@xiaokang xiaokang requested a review from dataroaring March 15, 2024 10:04
@doris-robot
Copy link

TPC-H: Total hot run time: 38577 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 93b3327eee13df906b5fa8105421cf1d4d9e692f, data reload: false

------ Round 1 ----------------------------------
q1	17643	4335	4164	4164
q2	2023	148	150	148
q3	10599	1055	893	893
q4	7434	739	709	709
q5	7464	2758	2746	2746
q6	185	125	122	122
q7	1181	837	828	828
q8	9347	2019	2011	2011
q9	7089	6480	6415	6415
q10	8542	3552	3652	3552
q11	429	232	222	222
q12	690	304	306	304
q13	17789	2870	2846	2846
q14	275	250	243	243
q15	506	457	444	444
q16	510	391	388	388
q17	958	507	586	507
q18	7219	6480	6565	6480
q19	3337	1417	1483	1417
q20	552	294	282	282
q21	6305	3549	3606	3549
q22	381	307	335	307
Total cold run time: 110458 ms
Total hot run time: 38577 ms

----- Round 2, with runtime_filter_mode=off -----
q1	4113	4077	4088	4077
q2	320	219	227	219
q3	2952	2853	2743	2743
q4	1828	1541	1528	1528
q5	5213	5245	5253	5245
q6	195	117	119	117
q7	2229	1867	1871	1867
q8	3172	3317	3301	3301
q9	8583	8546	8557	8546
q10	3675	3718	3670	3670
q11	537	433	444	433
q12	716	563	555	555
q13	16922	2851	2850	2850
q14	273	250	256	250
q15	480	436	442	436
q16	457	388	413	388
q17	1723	1476	1482	1476
q18	7600	7286	7193	7193
q19	1634	1525	1557	1525
q20	1903	1707	1712	1707
q21	4855	4557	4643	4557
q22	547	473	435	435
Total cold run time: 69927 ms
Total hot run time: 53118 ms

liaoxin01
liaoxin01 previously approved these changes Mar 16, 2024
Copy link
Contributor

@liaoxin01 liaoxin01 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 Mar 16, 2024
@github-actions
Copy link
Contributor

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

@github-actions
Copy link
Contributor

PR approved by anyone and no changes requested.

dataroaring
dataroaring previously approved these changes Mar 20, 2024
Copy link
Contributor

@dataroaring dataroaring left a comment

Choose a reason for hiding this comment

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

LGTM

@sollhui sollhui dismissed stale reviews from dataroaring and liaoxin01 via 9584ba3 March 20, 2024 15:48
@sollhui
Copy link
Contributor Author

sollhui commented Mar 20, 2024

run buildall

@github-actions github-actions bot removed the approved Indicates a PR has been approved by one committer. label Mar 20, 2024
@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@sollhui
Copy link
Contributor Author

sollhui commented Mar 21, 2024

run buildall

@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@sollhui
Copy link
Contributor Author

sollhui commented Mar 21, 2024

run buildall

@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@doris-robot
Copy link

TeamCity be ut coverage result:
Function Coverage: 35.26% (8728/24752)
Line Coverage: 27.08% (71460/263891)
Region Coverage: 26.31% (37064/140873)
Branch Coverage: 23.22% (18948/81618)
Coverage Report: http://coverage.selectdb-in.cc/coverage/ac4e726fc2a51378c32a3d03bd0a35f0b7cd9eec_ac4e726fc2a51378c32a3d03bd0a35f0b7cd9eec/report/index.html

liaoxin01
liaoxin01 previously approved these changes Mar 21, 2024
Copy link
Contributor

@liaoxin01 liaoxin01 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 Mar 21, 2024
@github-actions
Copy link
Contributor

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

@sollhui
Copy link
Contributor Author

sollhui commented Mar 21, 2024

run buildall

@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@doris-robot
Copy link

TPC-H: Total hot run time: 37695 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit e8c946b4fa62b719b8609c85e892f8a95cdc3ea2, data reload: false

------ Round 1 ----------------------------------
q1	17631	4320	4103	4103
q2	2020	154	143	143
q3	10595	1145	1184	1145
q4	10231	821	855	821
q5	7482	2992	2929	2929
q6	203	124	121	121
q7	1027	597	566	566
q8	9334	2017	1993	1993
q9	7125	6465	6444	6444
q10	8416	3341	3486	3341
q11	432	233	215	215
q12	404	201	205	201
q13	17786	2864	2835	2835
q14	228	198	214	198
q15	494	447	457	447
q16	454	360	350	350
q17	954	563	593	563
q18	7172	6632	6428	6428
q19	4230	1437	1386	1386
q20	522	248	249	248
q21	3519	3019	2925	2925
q22	359	293	319	293
Total cold run time: 110618 ms
Total hot run time: 37695 ms

----- Round 2, with runtime_filter_mode=off -----
q1	4114	4038	4066	4038
q2	319	222	216	216
q3	2964	2828	2807	2807
q4	1876	1534	1541	1534
q5	5219	5290	5253	5253
q6	193	115	119	115
q7	2255	1836	1854	1836
q8	3151	3263	3266	3263
q9	8540	8463	8564	8463
q10	3721	3660	3645	3645
q11	543	456	428	428
q12	713	551	545	545
q13	16919	2860	2862	2860
q14	272	256	255	255
q15	484	465	447	447
q16	463	424	423	423
q17	1756	1482	1479	1479
q18	7613	7159	7079	7079
q19	1586	1483	1567	1483
q20	1898	1714	1743	1714
q21	4848	4606	4877	4606
q22	509	412	440	412
Total cold run time: 69956 ms
Total hot run time: 52901 ms

@doris-robot
Copy link

TeamCity be ut coverage result:
Function Coverage: 35.27% (8730/24750)
Line Coverage: 27.09% (71494/263883)
Region Coverage: 26.33% (37085/140873)
Branch Coverage: 23.23% (18959/81620)
Coverage Report: http://coverage.selectdb-in.cc/coverage/e8c946b4fa62b719b8609c85e892f8a95cdc3ea2_e8c946b4fa62b719b8609c85e892f8a95cdc3ea2/report/index.html

@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@doris-robot
Copy link

TPC-H: Total hot run time: 37715 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 288f15c26f45787b14ba5af86f1b3d7a03d3df20, data reload: false

------ Round 1 ----------------------------------
q1	17682	4882	4198	4198
q2	2034	159	148	148
q3	10653	1162	1149	1149
q4	10227	788	784	784
q5	7456	3001	3003	3001
q6	200	123	124	123
q7	1032	599	544	544
q8	9330	1999	1987	1987
q9	7092	6521	6437	6437
q10	8396	3279	3465	3279
q11	442	219	226	219
q12	398	195	192	192
q13	17799	2849	2878	2849
q14	235	197	209	197
q15	507	455	452	452
q16	463	359	358	358
q17	976	575	544	544
q18	7350	6466	6456	6456
q19	1881	1404	1450	1404
q20	536	269	270	269
q21	3501	2821	3029	2821
q22	345	321	304	304
Total cold run time: 108535 ms
Total hot run time: 37715 ms

----- Round 2, with runtime_filter_mode=off -----
q1	4111	4084	4052	4052
q2	321	229	236	229
q3	2991	2886	2869	2869
q4	1826	1568	1560	1560
q5	5248	5311	5269	5269
q6	191	117	119	117
q7	2223	1852	1856	1852
q8	3149	3269	3289	3269
q9	8615	8619	8619	8619
q10	3733	3720	3669	3669
q11	563	441	442	441
q12	735	575	525	525
q13	16936	2842	2856	2842
q14	275	268	271	268
q15	488	442	445	442
q16	476	426	389	389
q17	1728	1472	1449	1449
q18	7563	7210	7250	7210
q19	1635	1507	1489	1489
q20	1899	1720	1688	1688
q21	4806	4778	4667	4667
q22	533	453	449	449
Total cold run time: 70045 ms
Total hot run time: 53364 ms

@doris-robot
Copy link

TPC-DS: Total hot run time: 181044 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 288f15c26f45787b14ba5af86f1b3d7a03d3df20, data reload: false

query1	938	353	350	350
query2	7381	2011	1969	1969
query3	6711	213	225	213
query4	31296	20662	20737	20662
query5	4308	403	402	402
query6	264	188	173	173
query7	4627	285	289	285
query8	226	179	165	165
query9	9964	2292	2264	2264
query10	461	236	248	236
query11	17059	14274	14476	14274
query12	127	87	84	84
query13	1650	410	411	410
query14	11080	10377	11022	10377
query15	249	212	192	192
query16	7902	267	272	267
query17	1919	575	549	549
query18	1986	295	275	275
query19	211	153	153	153
query20	95	89	83	83
query21	199	129	128	128
query22	4673	4460	4449	4449
query23	31934	30965	31330	30965
query24	10494	2857	2855	2855
query25	593	381	378	378
query26	715	157	155	155
query27	2340	346	350	346
query28	6277	1856	1878	1856
query29	866	644	627	627
query30	307	149	147	147
query31	970	742	736	736
query32	98	72	56	56
query33	665	262	250	250
query34	885	462	485	462
query35	829	620	610	610
query36	991	868	845	845
query37	118	76	78	76
query38	3602	3446	3441	3441
query39	1451	1368	1361	1361
query40	215	115	112	112
query41	52	48	46	46
query42	106	96	101	96
query43	493	464	439	439
query44	1071	718	705	705
query45	264	273	260	260
query46	1090	688	687	687
query47	1693	1627	1593	1593
query48	439	357	352	352
query49	1085	328	318	318
query50	777	365	368	365
query51	6600	6676	6606	6606
query52	101	86	88	86
query53	336	273	269	269
query54	322	227	244	227
query55	81	79	80	79
query56	252	219	225	219
query57	1093	1017	1030	1017
query58	221	194	199	194
query59	2736	2772	2613	2613
query60	257	234	247	234
query61	95	93	93	93
query62	606	428	382	382
query63	291	267	267	267
query64	5269	3897	3639	3639
query65	3104	2990	3006	2990
query66	864	345	333	333
query67	14875	14804	14505	14505
query68	5535	507	518	507
query69	583	380	391	380
query70	1187	1169	1189	1169
query71	443	274	277	274
query72	6390	2667	2591	2591
query73	723	320	322	320
query74	7860	6605	6605	6605
query75	3823	2947	2932	2932
query76	4148	842	1023	842
query77	641	265	262	262
query78	10423	9580	9709	9580
query79	8125	513	510	510
query80	1965	421	408	408
query81	557	214	213	213
query82	1056	198	214	198
query83	317	157	143	143
query84	290	82	80	80
query85	1669	389	371	371
query86	501	277	312	277
query87	3680	3510	3573	3510
query88	4947	2318	2299	2299
query89	477	355	361	355
query90	1950	175	169	169
query91	159	130	131	130
query92	60	46	46	46
query93	6151	501	476	476
query94	1222	172	172	172
query95	430	328	322	322
query96	615	263	264	263
query97	3076	2898	2843	2843
query98	214	208	201	201
query99	1071	771	744	744
Total cold run time: 301137 ms
Total hot run time: 181044 ms

@doris-robot
Copy link

TeamCity be ut coverage result:
Function Coverage: 35.28% (8734/24754)
Line Coverage: 27.09% (71510/264005)
Region Coverage: 26.33% (37103/140910)
Branch Coverage: 23.23% (18974/81678)
Coverage Report: http://coverage.selectdb-in.cc/coverage/288f15c26f45787b14ba5af86f1b3d7a03d3df20_288f15c26f45787b14ba5af86f1b3d7a03d3df20/report/index.html

@doris-robot
Copy link

Load test result on machine: 'aliyun_ecs.c7a.8xlarge_32C64G'

Load test result on commit 288f15c26f45787b14ba5af86f1b3d7a03d3df20 with default session variables
Stream load json:         19 seconds loaded 2358488459 Bytes, about 118 MB/s
Stream load orc:          59 seconds loaded 1101869774 Bytes, about 17 MB/s
Stream load parquet:      32 seconds loaded 861443392 Bytes, about 25 MB/s
Insert into select:       22.3 seconds inserted 10000000 Rows, about 448K ops/s

@sollhui
Copy link
Contributor Author

sollhui commented Mar 22, 2024

run buildall

@sollhui sollhui changed the title [opt](routine-load) optimize routine load task limit [opt](routine-load) optimize routine load task thread pool and related param Mar 22, 2024
@github-actions
Copy link
Contributor

clang-tidy review says "All clean, LGTM! 👍"

@doris-robot
Copy link

TPC-H: Total hot run time: 37614 ms
machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
Tpch sf100 test result on commit 9fff08c765daa9948d2060816543c6c9f8574c23, data reload: false

------ Round 1 ----------------------------------
q1	17621	4340	4128	4128
q2	2026	167	147	147
q3	10575	1148	1189	1148
q4	10234	742	787	742
q5	7476	2950	2973	2950
q6	199	124	124	124
q7	1021	590	562	562
q8	9342	2004	1964	1964
q9	7301	6479	6463	6463
q10	8389	3361	3465	3361
q11	433	232	224	224
q12	395	200	198	198
q13	17795	2843	2882	2843
q14	246	201	209	201
q15	517	459	476	459
q16	487	360	367	360
q17	954	622	549	549
q18	7219	6461	6366	6366
q19	2987	1436	1457	1436
q20	548	254	246	246
q21	3531	2891	2854	2854
q22	355	289	298	289
Total cold run time: 109651 ms
Total hot run time: 37614 ms

----- Round 2, with runtime_filter_mode=off -----
q1	4178	4079	4080	4079
q2	334	228	227	227
q3	2997	2883	2805	2805
q4	1847	1508	1562	1508
q5	5225	5263	5238	5238
q6	196	118	117	117
q7	2230	1831	1861	1831
q8	3164	3295	3303	3295
q9	8595	8582	8559	8559
q10	3707	3715	3707	3707
q11	555	449	443	443
q12	722	532	564	532
q13	16955	2825	2870	2825
q14	277	247	238	238
q15	497	465	453	453
q16	482	428	408	408
q17	1759	1490	1495	1490
q18	7535	7070	7022	7022
q19	1616	1528	1503	1503
q20	1915	1746	1709	1709
q21	4837	4672	4756	4672
q22	538	471	412	412
Total cold run time: 70161 ms
Total hot run time: 53073 ms

@doris-robot
Copy link

TPC-DS: Total hot run time: 181049 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 9fff08c765daa9948d2060816543c6c9f8574c23, data reload: false

query1	958	367	352	352
query2	7424	2047	1994	1994
query3	6706	208	201	201
query4	30969	20662	20676	20662
query5	4301	443	410	410
query6	275	185	169	169
query7	4631	292	292	292
query8	221	169	172	169
query9	9194	2283	2280	2280
query10	456	242	248	242
query11	17103	14232	14194	14194
query12	140	91	86	86
query13	1632	431	420	420
query14	11339	11073	10641	10641
query15	255	218	195	195
query16	8115	261	248	248
query17	1948	557	531	531
query18	2060	283	269	269
query19	226	148	153	148
query20	96	92	89	89
query21	204	126	124	124
query22	4608	4393	4407	4393
query23	31675	30931	30775	30775
query24	10614	2889	2850	2850
query25	611	384	375	375
query26	1200	154	159	154
query27	2656	353	351	351
query28	7338	1912	1886	1886
query29	872	645	634	634
query30	305	153	150	150
query31	949	735	736	735
query32	92	58	57	57
query33	782	263	254	254
query34	1082	507	492	492
query35	847	607	609	607
query36	995	876	846	846
query37	123	80	78	78
query38	3588	3458	3443	3443
query39	1535	1370	1508	1370
query40	215	116	113	113
query41	53	52	47	47
query42	104	98	96	96
query43	498	468	463	463
query44	1160	732	726	726
query45	281	271	276	271
query46	1099	713	700	700
query47	1657	1586	1594	1586
query48	438	352	354	352
query49	1136	343	349	343
query50	776	370	376	370
query51	6795	6603	6557	6557
query52	112	101	90	90
query53	351	278	278	278
query54	312	253	245	245
query55	89	80	84	80
query56	254	238	241	238
query57	1081	1003	1022	1003
query58	232	217	212	212
query59	2884	2601	2632	2601
query60	272	258	272	258
query61	119	113	111	111
query62	607	417	401	401
query63	307	280	277	277
query64	5783	3949	3938	3938
query65	3096	2981	2986	2981
query66	871	365	360	360
query67	14848	14378	14292	14292
query68	6910	522	523	522
query69	617	407	389	389
query70	1202	1139	1112	1112
query71	487	291	303	291
query72	6350	2672	2509	2509
query73	723	313	314	313
query74	7043	6568	6637	6568
query75	3939	2811	2754	2754
query76	4457	891	895	891
query77	626	258	258	258
query78	10376	9621	9532	9532
query79	12492	529	534	529
query80	1972	382	383	382
query81	537	219	211	211
query82	1333	190	206	190
query83	228	146	149	146
query84	291	78	79	78
query85	1498	326	318	318
query86	471	285	281	281
query87	3701	3500	3502	3500
query88	5441	2302	2304	2302
query89	534	379	367	367
query90	1998	180	177	177
query91	167	135	137	135
query92	65	49	48	48
query93	7547	501	493	493
query94	1251	182	179	179
query95	434	339	339	339
query96	594	276	270	270
query97	3043	2879	2848	2848
query98	240	215	201	201
query99	1100	761	772	761
Total cold run time: 311222 ms
Total hot run time: 181049 ms

@doris-robot
Copy link

Load test result on machine: 'aliyun_ecs.c7a.8xlarge_32C64G'

Load test result on commit 9fff08c765daa9948d2060816543c6c9f8574c23 with default session variables
Stream load json:         18 seconds loaded 2358488459 Bytes, about 124 MB/s
Stream load orc:          59 seconds loaded 1101869774 Bytes, about 17 MB/s
Stream load parquet:      31 seconds loaded 861443392 Bytes, about 26 MB/s
Insert into select:       21.9 seconds inserted 10000000 Rows, about 456K ops/s

@doris-robot
Copy link

TeamCity be ut coverage result:
Function Coverage: 35.28% (8734/24757)
Line Coverage: 27.08% (71518/264139)
Region Coverage: 26.32% (37107/140992)
Branch Coverage: 23.21% (18974/81738)
Coverage Report: http://coverage.selectdb-in.cc/coverage/9fff08c765daa9948d2060816543c6c9f8574c23_9fff08c765daa9948d2060816543c6c9f8574c23/report/index.html

@github-actions github-actions bot added the approved Indicates a PR has been approved by one committer. label Mar 22, 2024
@github-actions
Copy link
Contributor

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

Copy link
Contributor

@dataroaring dataroaring left a comment

Choose a reason for hiding this comment

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

LGTM

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. dev/2.0.13-merged dev/2.1.3-merged reviewed

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants