4.0查询出错 /tmp/1000_tidb/MC4wLjAuMDo0MDAwLzAuMC4wLjA6MTAwODA=/tmp-storage/chunk.ListInDisk566770649: no such file or directory

🪐 TiDB 技术问题 部署&运维管理
1

升级到4.0 rc2 之后 ，同样的查询，报错了：

1105 - open /tmp/1000_tidb/MC4wLjAuMDo0MDAwLzAuMC4wLjA6MTAwODA=/tmp-storage/chunk.ListInDisk566770649: no such file or directory

2
  1. 请问具体是什么查询，麻烦上传下，多谢。
  2. 是tidb.log 的报错吗？ 麻烦上传这段时间完整的tidb日志
3

碰到同样的问题：

1105 - open /tmp/1000_tidb/MC4wLjAuMDo0MDAwLzAuMC4wLjA6MTAwODA=/tmp-storage/chunk.ListInDisk909715224: no such file or directory, Time: 56.651000s

相关的 tidb.log tidb.log.zip (2.2 MB)

对了，我这边的 TiDB 版本是 5.7.25-TiDB-v4.0.0。

4

另外，tidb_stderr.log 里面应该没有相关的日志，因为它最后一条日志是 6月28日（两天前）。

5

对了，执行的是这个查询，

SELECT
user_sid,gjj_sid,apply_id,product_cid,apply_time,confirm_time,orderId
,inserttimespan,lc_countrate,lc_3countrate
,lc_6countrate,call_3time_15srate,top10_call_count
,top10_call_3count,top10_call_6time,call_6time,call_count,call_3count,call_type_countrate
,call_type_6countrate,total_3fee,now()
FROM
(
SELECT
a.apply_id
,a.user_sid
,a.gjj_sid
,a.product_cid
,a.apply_time
,a.confirm_time
,case when a.confirm_time > 0 then TIMESTAMPDIFF(minute,b.inserttimespan,a.confirm_time)
else TIMESTAMPDIFF(minute,b.inserttimespan,a.apply_time)end time_diff
,b.*
FROM dwd.dwd_gjj_apply_info a
LEFT JOIN hidden.hidden_mongo_tel_analysis_apply b on a.yys_order_sn = b.orderId
left join hidden.hidden_tel_analysis_apply c on a.apply_id=c.apply_id
WHERE b.orderId is not null and c.apply_id is null and a.confirm_time> 0
ORDER BY a.apply_id,abs(time_diff)
)a
GROUP BY apply_id;

其中 hidden.hidden_mongo_tel_analysis_apply 是一个视图，它的底表是有 json 字段的。

这是相应的执行计划，

 "Projection_18"	"2677251.24"	"root"		"dwd.dwd_gjj_apply_info.user_sid, dwd.dwd_gjj_apply_info.gjj_sid, dwd.dwd_gjj_apply_info.apply_id, dwd.dwd_gjj_apply_info.product_cid, dwd.dwd_gjj_apply_info.apply_time, dwd.dwd_gjj_apply_info.confirm_time, Column#52, Column#53, Column#54, Column#55, Column#56, Column#57, Column#58, Column#59, Column#60, Column#61, Column#62, Column#63, Column#64, Column#65, Column#66, 2020-06-30 16:22:04->Column#90"
"└─Projection_112"	"2677251.24"	"root"		"dwd.dwd_gjj_apply_info.apply_id, dwd.dwd_gjj_apply_info.user_sid, dwd.dwd_gjj_apply_info.gjj_sid, dwd.dwd_gjj_apply_info.product_cid, dwd.dwd_gjj_apply_info.apply_time, dwd.dwd_gjj_apply_info.confirm_time, Column#89, Column#52, Column#53, Column#54, Column#55, Column#56, Column#57, Column#58, Column#59, Column#60, Column#61, Column#62, Column#63, Column#64, Column#65, Column#66"
"  └─Sort_19"	"2677251.24"	"root"		"dwd.dwd_gjj_apply_info.apply_id:asc, Column#148:asc"
"    └─Projection_113"	"2677251.24"	"root"		"dwd.dwd_gjj_apply_info.apply_id, dwd.dwd_gjj_apply_info.user_sid, dwd.dwd_gjj_apply_info.gjj_sid, dwd.dwd_gjj_apply_info.product_cid, dwd.dwd_gjj_apply_info.apply_time, dwd.dwd_gjj_apply_info.confirm_time, Column#89, Column#52, Column#53, Column#54, Column#55, Column#56, Column#57, Column#58, Column#59, Column#60, Column#61, Column#62, Column#63, Column#64, Column#65, Column#66, abs(Column#89)->Column#148"
"      └─Projection_20"	"2677251.24"	"root"		"dwd.dwd_gjj_apply_info.apply_id, dwd.dwd_gjj_apply_info.user_sid, dwd.dwd_gjj_apply_info.gjj_sid, dwd.dwd_gjj_apply_info.product_cid, dwd.dwd_gjj_apply_info.apply_time, dwd.dwd_gjj_apply_info.confirm_time, case(gt(dwd.dwd_gjj_apply_info.confirm_time, 0000-00-00 00:00:00), timestampdiff(MINUTE, Column#53, dwd.dwd_gjj_apply_info.confirm_time), timestampdiff(MINUTE, Column#53, dwd.dwd_gjj_apply_info.apply_time))->Column#89, Column#52, Column#53, Column#54, Column#55, Column#56, Column#57, Column#58, Column#59, Column#60, Column#61, Column#62, Column#63, Column#64, Column#65, Column#66"
"        └─Selection_21"	"2677251.24"	"root"		"isnull(hidden.hidden_tel_analysis_apply.apply_id)"
"          └─HashJoin_32"	"3346564.05"	"root"		"left outer join, equal:[eq(dwd.dwd_gjj_apply_info.apply_id, hidden.hidden_tel_analysis_apply.apply_id)]"
"            ├─IndexReader_107(Build)"	"3922787.00"	"root"		"index:IndexFullScan_106"
"            │ └─IndexFullScan_106"	"3922787.00"	"cop[tikv]"	"table:c, index:idx_orderid(user_sid)"	"keep order:false"
"            └─HashJoin_50(Probe)"	"3346564.05"	"root"		"inner join, equal:[eq(dwd.dwd_gjj_apply_info.yys_order_sn, Column#52)]"
"              ├─TableReader_99(Build)"	"3269229.55"	"root"		"data:Selection_98"
"              │ └─Selection_98"	"3269229.55"	"cop[tikv]"		"gt(dwd.dwd_gjj_apply_info.confirm_time, 0000-00-00 00:00:00), not(isnull(dwd.dwd_gjj_apply_info.yys_order_sn))"
"              │   └─TableFullScan_97"	"10762000.00"	"cop[tikv]"	"table:a"	"keep order:false"
"              └─Projection_51(Probe)"	"1446548.74"	"root"		"mongo.mongo_mobile_report_data_recent_analysis.orderid, date_sub(str_to_date(json_unquote(cast(json_extract(mongo.mongo_mobile_datas.thedata, $.create_date), var_string(16777216))), %Y-%m-%d %h:%i:%s), 8, HOUR)->Column#53, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.lc_countrate), var_string(16777216)))->Column#54, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.lc_3countrate), var_string(16777216)))->Column#55, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.lc_6countrate), var_string(16777216)))->Column#56, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_3time_15srate), var_string(16777216)))->Column#57, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.top10_call_count), var_string(16777216)))->Column#58, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.top10_call_3count), var_string(16777216)))->Column#59, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.top10_call_6time), var_string(16777216)))->Column#60, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_6time), var_string(16777216)))->Column#61, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_count), var_string(16777216)))->Column#62, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_3count), var_string(16777216)))->Column#63, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_type_countrate), var_string(16777216)))->Column#64, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.call_type_6countrate), var_string(16777216)))->Column#65, json_unquote(cast(json_extract(mongo.mongo_mobile_report_data_recent_analysis.thedata, $.total_3fee), var_string(16777216)))->Column#66"
"                └─IndexMergeJoin_62"	"1446548.74"	"root"		"inner join, inner:Projection_60, outer key:mongo.mongo_mobile_report_data_recent_analysis.orderid, inner key:mongo.mongo_mobile_datas.orderid"
"                  ├─TableReader_91(Build)"	"1433986.00"	"root"		"data:Selection_90"
"                  │ └─Selection_90"	"1433986.00"	"cop[tikv]"		"not(isnull(mongo.mongo_mobile_report_data_recent_analysis.orderid))"
"                  │   └─TableFullScan_89"	"1433986.00"	"cop[tikv]"	"table:a"	"keep order:false"
"                  └─Projection_60(Probe)"	"1.00"	"root"		"mongo.mongo_mobile_datas.orderid, mongo.mongo_mobile_datas.thedata"
"                    └─IndexLookUp_59"	"1.00"	"root"		
"                      ├─IndexRangeScan_57(Build)"	"1.00"	"cop[tikv]"	"table:d, index:PRIMARY(orderid)"	"range: decided by [eq(mongo.mongo_mobile_datas.orderid, mongo.mongo_mobile_report_data_recent_analysis.orderid)], keep order:true"
"                      └─TableRowIDScan_58(Probe)"	"1.00"	"cop[tikv]"	"table:d"	"keep order:false"
6

我也遇到了，而且还遇到查询时间长了出现TiKV server timeout ，感觉4.0问题蛮大

7

@tminglei 下次可以新开个帖子，会比较及时回复。:grinning: 这个问题我们查一下。

8

是否方便反馈一下上面的信息，方便查询？

9

查询的执行计划：

Projection_16 10545732947.55 root div(test.stock.open, test.stock.close)->Column#41, div(test.stock.high, test.stock.close)->Column#42, div(test.stock.low, test.stock.close)->Column#43, div(test.stock.close, test.stock.close)->Column#44, div(test.stock.volume, test.stock.volume)->Column#45, div(test.stock.amount, test.stock.amount)->Column#46, test.stock.p_change, div(test.stock.open, test.stock.close)->Column#47, div(test.stock.high, test.stock.close)->Column#48, div(test.stock.low, test.stock.close)->Column#49, cast(plus(div(test.stock.p_change, 3), 4), bigint(22) UNSIGNED BINARY)->Column#50, test.stock.date, test.stock.p_change
└─HashJoin_34 10545732947.55 root inner join, equal:[eq(test.days.z_day, test.stock.date) eq(test.stock.code, test.stock.code)]
├─HashJoin_140(Build) 9167042.52 root inner join, equal:[eq(test.days.m_day, test.stock.date) eq(test.stock.code, test.stock.code)]
│ ├─HashJoin_201(Build) 7968.59 root inner join, equal:[eq(test.days.n_day, test.stock.date)]
│ │ ├─IndexLookUp_218(Build) 6.93 root
│ │ │ ├─IndexRangeScan_215(Build) 6.93 cop[tikv] table:d, index:day_index(n_day) range:(2020-01-01,2020-01-09), keep order:false
│ │ │ └─Selection_217(Probe) 6.93 cop[tikv] not(isnull(test.days.m_day)), not(isnull(test.days.z_day))
│ │ │ └─TableRowIDScan_216 6.93 cop[tikv] table:d keep order:false
│ │ └─IndexLookUp_240(Probe) 10976.92 root
│ │ ├─IndexRangeScan_237(Build) 10976.92 cop[tikv] table:d1, index:index_date(date) range:(2020-01-01,2020-01-09), keep order:false
│ │ └─Selection_239(Probe) 10976.92 cop[tikv] not(isnull(test.stock.code))
│ │ └─TableRowIDScan_238 10976.92 cop[tikv] table:d1 keep order:false
│ └─TableReader_246(Probe) 10288358.00 root data:Selection_245
│ └─Selection_245 10288358.00 cop[tiflash] not(isnull(test.stock.code)), not(isnull(test.stock.date))
│ └─TableFullScan_244 10288358.00 cop[tiflash] table:d2 keep order:false
└─TableReader_260(Probe) 10288358.00 root data:Selection_259
└─Selection_259 10288358.00 cop[tiflash] not(isnull(test.stock.code)), not(isnull(test.stock.date))
└─TableFullScan_258 10288358.00 cop[tiflash] table:d3 keep order:false
10

tidb日志：>

[2020/07/08 11:17:09.085 +08:00] [INFO] [region_cache.go:533] [“mark store’s regions need be refill”] [store=192.168.0.230:3930]
[2020/07/08 11:17:09.085 +08:00] [INFO] [region_cache.go:552] [“switch region peer to next due to send request fail”] [conn=13534] [current=“region ID: 72969, meta: id:72969 start_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\201\327\244" end_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\211h\252" region_epoch:<conf_ver:6 version:3566 > peers:<id:72970 store_id:1 > peers:<id:72971 store_id:4 > peers:<id:72972 store_id:7 > peers:<id:81113 store_id:81081 is_learner:true > , peer: id:81113 store_id:81081 is_learner:true , addr: 192.168.0.230:3930, idx: 3”] [needReload=false] [error=“context deadline exceeded”]
[2020/07/08 11:17:10.980 +08:00] [INFO] [region_cache.go:414] [“invalidate current region, because others failed on same store”] [region=72969] [store=192.168.0.230:3930]
[2020/07/08 11:17:10.981 +08:00] [INFO] [coprocessor.go:871] [“[TIME_COP_PROCESS] resp_time:9.823120844s txnStartTS:417904127911657479 region_id:72969 store_addr:192.168.0.230:3930 backoff_ms:41637 backoff_types:[regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC]”] [conn=13534]
[2020/07/08 11:17:10.981 +08:00] [WARN] [backoff.go:319] [“regionMiss backoffer.maxSleep 40000ms is exceeded, errors:
region_id:72969, region_ver:3566, store_type:tiflash, peer_addr:192.168.0.230:3930, error:epoch_not_match:<> at 2020-07-08T11:17:00.657322631+08:00
send tikv request error: context deadline exceeded, ctx: region ID: 72969, meta: id:72969 start_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\201\327\244" end_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\211h\252" region_epoch:<conf_ver:6 version:3566 > peers:<id:72970 store_id:1 > peers:<id:72971 store_id:4 > peers:<id:72972 store_id:7 > peers:<id:81113 store_id:81081 is_learner:true > , peer: id:81113 store_id:81081 is_learner:true , addr: 192.168.0.230:3930, idx: 3, try next peer later at 2020-07-08T11:17:09.085724095+08:00
region_id:72969, region_ver:3566, store_type:tiflash, peer_addr:192.168.0.230:3930, error:epoch_not_match:<> at 2020-07-08T11:17:10.981104026+08:00”]
[2020/07/08 11:17:14.085 +08:00] [WARN] [client_batch.go:223] [“init create streaming fail”] [target=192.168.0.230:3930] [error=“context deadline exceeded”]
[2020/07/08 11:17:14.085 +08:00] [INFO] [region_cache.go:1523] [“[liveness] request kv status fail”] [store=192.168.0.230:20292] [error=“Get http://192.168.0.230:20292/status: dial tcp 192.168.0.230:20292: connect: no route to host”]
[2020/07/08 11:17:14.085 +08:00] [INFO] [region_cache.go:552] [“switch region peer to next due to send request fail”] [conn=13534] [current=“region ID: 72965, meta: id:72965 start_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000zF\236" end_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\201\327\244" region_epoch:<conf_ver:6 version:3565 > peers:<id:72966 store_id:1 > peers:<id:72967 store_id:4 > peers:<id:72968 store_id:7 > peers:<id:81112 store_id:81081 is_learner:true > , peer: id:81112 store_id:81081 is_learner:true , addr: 192.168.0.230:3930, idx: 3”] [needReload=false] [error=“context deadline exceeded”]
[2020/07/08 11:17:15.482 +08:00] [INFO] [region_cache.go:414] [“invalidate current region, because others failed on same store”] [region=72965] [store=192.168.0.230:3930]
[2020/07/08 11:17:15.482 +08:00] [INFO] [coprocessor.go:871] [“[TIME_COP_PROCESS] resp_time:9.551379169s txnStartTS:417904127911657479 region_id:72965 store_addr:192.168.0.230:3930 backoff_ms:40506 backoff_types:[regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC,regionMiss,tikvRPC]”] [conn=13534]
[2020/07/08 11:17:15.482 +08:00] [WARN] [backoff.go:319] [“regionMiss backoffer.maxSleep 40000ms is exceeded, errors:
region_id:72965, region_ver:3565, store_type:tiflash, peer_addr:192.168.0.230:3930, error:epoch_not_match:<> at 2020-07-08T11:17:05.429825937+08:00
send tikv request error: context deadline exceeded, ctx: region ID: 72965, meta: id:72965 start_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000zF\236" end_key:"t\200\000\000\000\000\000#\377_r\200\000\000\000\000\201\327\244" region_epoch:<conf_ver:6 version:3565 > peers:<id:72966 store_id:1 > peers:<id:72967 store_id:4 > peers:<id:72968 store_id:7 > peers:<id:81112 store_id:81081 is_learner:true > , peer: id:81112 store_id:81081 is_learner:true , addr: 192.168.0.230:3930, idx: 3, try next peer later at 2020-07-08T11:17:14.086006556+08:00
region_id:72965, region_ver:3565, store_type:tiflash, peer_addr:192.168.0.230:3930, error:epoch_not_match:<> at 2020-07-08T11:17:15.482461952+08:00”]
[2020/07/08 11:17:15.483 +08:00] [ERROR] [conn.go:728] [“command dispatched failed”] [conn=13534] [connInfo=“id:13534, addr:192.168.0.250:55748 status:10, collation:utf8_general_ci, user:root”] [command=Query] [status=“inTxn:0, autocommit:1”] [sql=“select d1.open/d3.close,d1.high/d3.close,d1.low/d3.close,d1.close/d3.close,d1.volume/d3.volume,d1.amount/d3.amount,\r
d3.p_change,d3.open/d3.close,d3.high/d3.close,d3.low/d3.close, cast(d2.p_change/3+4 as UNSIGNED)\r
,d1.date,d2.p_change\r
from stock d1\r
left join days d on d1.date=d.n_day\r
inner join stock d2 on d2.date=d.m_day and d2.code=d1.code\r
inner join stock d3 on d3.date=d.z_day and d3.code=d1.code\r
\t\t\t\t\t\twhere d1.date>‘2020-01-01’ and d1.date<‘2020-01-09’”] [txn_mode=OPTIMISTIC] [err=“[tikv:9002]TiKV server timeout
github.com/pingcap/errors.AddStack
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/pkg/mod/github.com/pingcap/errors@v0.11.5-0.20190809092503-95897b64e011/errors.go:174
github.com/pingcap/errors.Trace
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/pkg/mod/github.com/pingcap/errors@v0.11.5-0.20190809092503-95897b64e011/juju_adaptor.go:15
github.com/pingcap/tidb/store/tikv.(*copIteratorWorker).handleCopResponse
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/src/github.com/pingcap/tidb/store/tikv/coprocessor.go:934
github.com/pingcap/tidb/store/tikv.(*copIteratorWorker).handleTaskOnce
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/src/github.com/pingcap/tidb/store/tikv/coprocessor.go:758
github.com/pingcap/tidb/store/tikv.(*copIteratorWorker).handleTask
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/src/github.com/pingcap/tidb/store/tikv/coprocessor.go:672
github.com/pingcap/tidb/store/tikv.(*copIteratorWorker).run
\t/home/jenkins/agent/workspace/tidb_v4.0.0-rc.2/go/src/github.com/pingcap/tidb/store/tikv/coprocessor.go:494
runtime.goexit
\t/usr/local/go/src/runtime/asm_amd64.s:1357”]

11

感谢，已经在查看问题了，有进展会尽快答复，多谢。

12

好的，辛苦了~

14

hi， 是每次必出现还是偶发的？

15

hi，麻烦确认一下

  1. 临时文件目录是否存在…
  2. 是否稳定复现这个问题。
  3. 在已经出现这个问题的 TiDB 上，将 tidb_mem_quota_query 调足够小，执行2表 Join并落盘，看看 TiDB 在这个时候能否正常工作。
16

两个问题都是在执行我发的SQL必然触发的

17

1，目录是存在的

image

2，稳定出现这个问题

3，我没有调过tidb_mem_quota_query参数，不确定是不是这个造成的，但是我查询的表的确有点大，大概900w的数据表

18

好的，我们确认一下问题，有进度及时同步，感谢理解

19

感谢你的回复 目前还有几个问题需要确认

  1. 集群是从那个版本升级到 4.0 rc2 的？ tidb_mem_quota_query 这个参数是 4.0 之后引入的参数，表示对单条 SQL 使用内存大小的限制，默认为 1 GB。
  2. 集群的拓扑是怎样的，有多台 TiDB 吗？多个 TiDB 示例是否有部署到同一台机器上？是否每个 TiDB 都能稳定复现这个问题
  3. 为了排查问题，麻烦新开一个session，set @@session.tidb_mem_quota_query = 1; 将内存限制调整到1b，然后执行一条会落盘的语句，如 select * from d order by xxxx, 看看是否能正常触发落盘
  4. 如果机器内存够用，无需尝试落盘，可以把 config 文件中的 mem-quota-query 设置成 4.0-版本之前默认的 32G，oom-use-tmp-storage设置成 false，来关闭中间数据落盘的功能。
20

1：版本为：5.7.25-TiDB-v4.0.0-rc.2，默认的确是1G
2：目前只有一个tidb节点，三个tikv和一个tiflash
3：
image
4：我改一下看看

21

@Mingdr 您好，请问，有测试结论了吗？多谢。