You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@doris.apache.org by "luwei16 (via GitHub)" <gi...@apache.org> on 2026/01/08 17:01:21 UTC

[PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

luwei16 opened a new pull request, #59698:
URL: https://github.com/apache/doris/pull/59698

   
   
   Add address-provider retry path in ThriftRpcHelper; stream load uses provider to retry on new FE master Only reconnect MetaServiceProxy on request failures; close connection on getVisibleVersionAsync failure Update meta_service_rpc_reconnect_interval_ms in FE/BE (5000 -> 100)


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "hello-stephen (via GitHub)" <gi...@apache.org>.
hello-stephen commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3725052617

   # FE UT Coverage Report
   Increment line coverage `  59.09% (13/22)` :tada:
   [Increment coverage report](http://coverage.selectdb-in.cc/coverage/59698_310a60f02210f32e9ed54e26c39f5bb2720690f8/fe_increment_report/index.html)
   [Complete coverage report](http://coverage.selectdb-in.cc/coverage/59698_310a60f02210f32e9ed54e26c39f5bb2720690f8/fe_report/index.html)
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "doris-robot (via GitHub)" <gi...@apache.org>.
doris-robot commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3727142505

   
   <details>
   <summary>TPC-H: <b>Total hot run time: 31600 ms</b></summary>
   
   ```
   machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
   scripts: https://github.com/apache/doris/tree/master/tools/tpch-tools
   Tpch sf100 test result on commit 0cfce02a7c70c3795baeb4f305199b632db80471, data reload: false
   
   ------ Round 1 ----------------------------------
   q1	17629	4160	4086	4086
   q2	2084	362	247	247
   q3	10136	1313	702	702
   q4	10198	846	321	321
   q5	7504	2035	1842	1842
   q6	193	175	139	139
   q7	928	791	649	649
   q8	9276	1344	1158	1158
   q9	5054	4542	4563	4542
   q10	6803	1818	1414	1414
   q11	528	292	273	273
   q12	695	752	617	617
   q13	17787	3880	3111	3111
   q14	295	293	267	267
   q15	602	511	506	506
   q16	713	681	632	632
   q17	675	796	519	519
   q18	6929	6492	6522	6492
   q19	1100	984	592	592
   q20	397	361	251	251
   q21	2999	2424	2266	2266
   q22	1034	1002	974	974
   Total cold run time: 103559 ms
   Total hot run time: 31600 ms
   
   ----- Round 2, with runtime_filter_mode=off -----
   q1	4091	4057	4041	4041
   q2	336	382	311	311
   q3	2081	2599	2173	2173
   q4	1328	1754	1317	1317
   q5	4117	4004	4052	4004
   q6	218	173	133	133
   q7	1900	1796	1698	1698
   q8	2864	2545	2428	2428
   q9	7397	7145	7161	7145
   q10	2473	2788	2285	2285
   q11	554	481	452	452
   q12	728	746	610	610
   q13	3614	4114	3368	3368
   q14	299	319	275	275
   q15	662	615	611	611
   q16	731	683	619	619
   q17	1159	1292	1362	1292
   q18	8243	7984	7692	7692
   q19	996	841	868	841
   q20	1958	2068	1878	1878
   q21	4836	4540	4432	4432
   q22	1145	1079	1012	1012
   Total cold run time: 51730 ms
   Total hot run time: 48617 ms
   ```
   </details>
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "hello-stephen (via GitHub)" <gi...@apache.org>.
hello-stephen commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3724807491

   
   Thank you for your contribution to Apache Doris.
   Don't know what should be done next? See [How to process your PR](https://cwiki.apache.org/confluence/display/DORIS/How+to+process+your+PR).
   
   Please clearly describe your PR:
   1. What problem was fixed (it's best to include specific error reporting information). How it was fixed.
   2. Which behaviors were modified. What was the previous behavior, what is it now, why was it modified, and what possible impacts might there be.
   3. What features were added. Why was this function added?
   4. Which code was refactored and why was this part of the code refactored?
   5. Which functions were optimized and what is the difference before and after the optimization?
       


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "hello-stephen (via GitHub)" <gi...@apache.org>.
hello-stephen commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3738239520

   skip check_coverage


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "gavinchou (via GitHub)" <gi...@apache.org>.
gavinchou commented on code in PR #59698:
URL: https://github.com/apache/doris/pull/59698#discussion_r2673179835


##########
be/src/util/thrift_rpc_helper.cpp:
##########
@@ -85,15 +94,36 @@ Status ThriftRpcHelper::rpc(const std::string& ip, const int32_t port,
 #endif
             std::this_thread::sleep_for(
                     std::chrono::milliseconds(config::thrift_client_retry_interval_ms));
-            status = client.reopen(timeout_ms);
-            if (!status.ok()) {
+            TNetworkAddress retry_address = address_provider();
+            if (retry_address.hostname.empty() || retry_address.port == 0) {
+                return Status::Error<ErrorCode::SERVICE_UNAVAILABLE>("FE address is not available");
+            }
+            if (retry_address.hostname != address.hostname || retry_address.port != address.port) {
+#ifndef ADDRESS_SANITIZER
+                LOG(INFO) << "retrying call frontend service with new address=" << retry_address;

Review Comment:
   why not just log without check `ifndef address`? 



##########
be/src/util/thrift_rpc_helper.cpp:
##########
@@ -85,15 +94,36 @@ Status ThriftRpcHelper::rpc(const std::string& ip, const int32_t port,
 #endif
             std::this_thread::sleep_for(
                     std::chrono::milliseconds(config::thrift_client_retry_interval_ms));
-            status = client.reopen(timeout_ms);
-            if (!status.ok()) {
+            TNetworkAddress retry_address = address_provider();
+            if (retry_address.hostname.empty() || retry_address.port == 0) {
+                return Status::Error<ErrorCode::SERVICE_UNAVAILABLE>("FE address is not available");
+            }
+            if (retry_address.hostname != address.hostname || retry_address.port != address.port) {
+#ifndef ADDRESS_SANITIZER
+                LOG(INFO) << "retrying call frontend service with new address=" << retry_address;
+#endif
+                Status retry_status;
+                ClientConnection<T> retry_client(_s_exec_env->get_client_cache<T>(), retry_address,
+                                                 timeout_ms, &retry_status);
+                if (!retry_status.ok()) {
+#ifndef ADDRESS_SANITIZER
+                    LOG(WARNING) << "Connect frontend failed, address=" << retry_address

Review Comment:
   ditto



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3724856622

   PR approved by anyone and no changes requested.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3724856464

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


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "github-actions[bot] (via GitHub)" <gi...@apache.org>.
github-actions[bot] commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3729084516

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


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "luwei16 (via GitHub)" <gi...@apache.org>.
luwei16 commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3724815333

   run buildall


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "luwei16 (via GitHub)" <gi...@apache.org>.
luwei16 merged PR #59698:
URL: https://github.com/apache/doris/pull/59698


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "luwei16 (via GitHub)" <gi...@apache.org>.
luwei16 commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3726833212

   run buildall


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org


Re: [PR] [enhancement](cloud) improve FE RPC retry and MetaService connection handling [doris]

Posted by "doris-robot (via GitHub)" <gi...@apache.org>.
doris-robot commented on PR #59698:
URL: https://github.com/apache/doris/pull/59698#issuecomment-3727160657

   
   <details>
   <summary>TPC-DS: <b>Total hot run time: 173604 ms</b></summary>
   
   ```
   machine: 'aliyun_ecs.c7a.8xlarge_32C64G'
   scripts: https://github.com/apache/doris/tree/master/tools/tpcds-tools
   TPC-DS sf100 test result on commit 0cfce02a7c70c3795baeb4f305199b632db80471, data reload: false
   
   query5	4389	593	423	423
   query6	327	226	213	213
   query7	4219	458	280	280
   query8	359	255	242	242
   query9	8752	2697	2711	2697
   query10	497	391	303	303
   query11	15261	15292	14749	14749
   query12	177	114	116	114
   query13	1258	492	388	388
   query14	5927	2986	2744	2744
   query14_1	2682	2650	2639	2639
   query15	203	197	175	175
   query16	990	488	468	468
   query17	1128	673	549	549
   query18	2433	420	327	327
   query19	219	216	190	190
   query20	123	115	114	114
   query21	225	156	127	127
   query22	4033	4093	3920	3920
   query23	16126	15650	15421	15421
   query23_1	15510	15629	15449	15449
   query24	7451	1523	1175	1175
   query24_1	1167	1173	1180	1173
   query25	543	445	401	401
   query26	1232	260	153	153
   query27	2777	444	285	285
   query28	4593	2162	2128	2128
   query29	799	526	442	442
   query30	311	229	213	213
   query31	807	642	548	548
   query32	80	67	64	64
   query33	534	336	291	291
   query34	897	885	533	533
   query35	726	761	685	685
   query36	885	904	798	798
   query37	131	93	79	79
   query38	2805	2728	2626	2626
   query39	788	762	737	737
   query39_1	712	747	717	717
   query40	226	138	122	122
   query41	75	68	66	66
   query42	108	105	104	104
   query43	517	442	431	431
   query44	1303	730	736	730
   query45	192	187	181	181
   query46	849	968	608	608
   query47	1418	1444	1388	1388
   query48	327	329	260	260
   query49	621	428	334	334
   query50	643	273	200	200
   query51	3713	3908	3748	3748
   query52	108	109	96	96
   query53	300	329	278	278
   query54	298	289	257	257
   query55	79	78	74	74
   query56	306	303	298	298
   query57	1025	1040	950	950
   query58	282	262	259	259
   query59	2165	2162	2012	2012
   query60	327	340	317	317
   query61	197	190	190	190
   query62	400	366	324	324
   query63	299	268	272	268
   query64	5068	1301	1002	1002
   query65	3862	3757	3736	3736
   query66	1464	410	310	310
   query67	15309	15696	15561	15561
   query68	5605	991	716	716
   query69	489	340	302	302
   query70	1034	941	957	941
   query71	365	303	274	274
   query72	6154	3480	3422	3422
   query73	756	716	306	306
   query74	8784	8751	8618	8618
   query75	2828	2817	2425	2425
   query76	3870	1060	641	641
   query77	518	383	280	280
   query78	9821	10001	9151	9151
   query79	1553	908	580	580
   query80	749	579	470	470
   query81	535	267	233	233
   query82	413	145	110	110
   query83	267	253	237	237
   query84	265	110	100	100
   query85	923	531	475	475
   query86	360	326	313	313
   query87	2856	2890	2734	2734
   query88	3201	2231	2212	2212
   query89	395	353	336	336
   query90	2017	157	152	152
   query91	180	173	143	143
   query92	68	68	65	65
   query93	1367	870	529	529
   query94	622	318	301	301
   query95	591	328	303	303
   query96	573	458	201	201
   query97	2345	2376	2360	2360
   query98	219	199	190	190
   query99	612	570	513	513
   Total cold run time: 253158 ms
   Total hot run time: 173604 ms
   ```
   </details>
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org