flink sql 不是处于 RUNNING 状态而是 FINISHED 状态中止检查点

如何解决flink sql 不是处于 RUNNING 状态而是 FINISHED 状态中止检查点

启动Flink SQL进行数据处理时，发现checkpoint异常：

2021-06-16 14:27:57,484 INFO  org.apache.flink.runtime.checkpoint.CheckpointCoordinator    [] - Checkpoint triggering task Source: TableSourceScan(table=[[default_catalog,default_database,DIC_REGION,project=[PROVINCEID,PROVINCENAME]]],fields=[PROVINCEID,PROVINCENAME]) -> Calc(select=[(PROVINCEID SUBSTR 1 SUBSTR 2) AS PROVINCEID,PROVINCENAME]) (1/1) of job 786df250b9367ce2f20409578acf5ef5 is not in state RUNNING but FINISHED instead. Aborting checkpoint.
2021-06-16 14:28:07,PROVINCENAME]) (1/1) of job 786df250b9367ce2f20409578acf5ef5 is not in state RUNNING but FINISHED instead. Aborting checkpoint.

查看Webui时，发现我的程序代码表相关部分确实存在，状态为finished。

flink version:flink-1.12.0

flink sql:

INSERT INTO DM_D_BRIDGE_INFO_RT
SELECT
   TMP.LOANID,'NOLOANACNO' LOANACNO,A.CUSTID,A.PRODID,A.PROJNUM,A.DEPTID,A.CUSTNAME,A.AREAKIND,A.BUSINO,A.CHANNELNO,IF(A.ISFUNDFLAG='1','yes','no') ISFUNDFLAG,IF(IF(A.BUSI_TYPE is null,'1',A.BUSI_TYPE)=2,'applFisrt','applSecond') BUSI_TYPE,A.BUSI_CREDID CREDID,IF(A.DISCOUNTNUM is null,'no','yes') ISDISCOUNT,A.DISCOUNTNUM,A.DISCOUNTNAME,IF(A.TIEFEINUM is null,'yes') ISFEEPERC,A.TIEFEINUM,A.TIEFEINAME,A.WAYKIND INDUSTRYID,IF(A.ISCONTLOAN='1','no') ISCONTLOAN,A.TAPEOPERID MANAGERID_FKQ,'yes') ISDISCOUNT1,'yes') ISFEEPERC1,A.TTERM STTERM,IF(B1.PROVINCENAME is null,'no Arae Province',B1.PROVINCENAME) PROVINCE,IF(B2.CITYNAME is null,'no Arae City',B2.CITYNAME) CITY,IF(B3.DISTRICTNAME is null,'no Arae District',B3.DISTRICTNAME) DISTRICTNAME,C1.VALUENAME APPRSTATE,C2.VALUENAME LOANOBJECTSIZE,IF(E1.PROVINCENAME is null,'no Arae Channel',E1.PROVINCENAME) PROVINCE_CHANNEL,IF(E2.CITYNAME is null,E2.CITYNAME) CITY_CHANNEL,IF(E3.DISTRICTNAME is null,E3.DISTRICTNAME) DISTRICTNAME_CHANNEL,B2.CITYTYPE,IF(F.REGIONNAME is null,'no arae',F.REGIONNAME) REGIONS
 FROM TMP_XD_BUSI_APP AS TMP
   LEFT OUTER JOIN XD_BUSI_APP FOR SYSTEM_TIME AS OF TMP.PROCTIME AS A ON TMP.LOANID = A.LOANID
   LEFT OUTER JOIN (SELECT DISTINCT SUBSTR(PROVINCEID,1,2) PROVINCEID,PROVINCENAME FROM DIC_REGION) AS B1 ON B1.PROVINCEID=SUBSTR(A.AREAKIND,2)
   LEFT OUTER JOIN (SELECT DISTINCT SUBSTR(CITYID,4) CITYID,CITYNAME,CITYTYPE FROM DIC_REGION) AS B2 ON B2.CITYID=SUBSTR(A.AREAKIND,4)
   LEFT OUTER JOIN DIC_REGION FOR SYSTEM_TIME AS OF TMP.PROCTIME AS B3 ON B3.DISTRICTID=A.AREAKIND
   LEFT OUTER JOIN DIC_CODEVALUE FOR SYSTEM_TIME AS OF TMP.PROCTIME AS C1 ON C1.CODETYPE='apprstate' AND C1.VALUEID=A.APPRSTATE
   LEFT OUTER JOIN DIC_CODEVALUE FOR SYSTEM_TIME AS OF TMP.PROCTIME AS C2 ON C2.CODETYPE='loanobjectsize' AND C2.VALUEID=A.LOANOBJECTSIZE
   LEFT OUTER JOIN XD_SYS_PARA_CHANNEL FOR SYSTEM_TIME AS OF TMP.PROCTIME AS D ON D.CHANNELNO =A.CHANNELNO
   LEFT OUTER JOIN (SELECT DISTINCT PROVINCEID,PROVINCENAME FROM DIC_REGION) AS E1 ON D.PROVINCE=E1.PROVINCEID
   LEFT OUTER JOIN (SELECT DISTINCT CITYID,CITYNAME FROM DIC_REGION) AS E2 ON D.CITY=E2.CITYID
   LEFT OUTER JOIN (SELECT DISTINCT DISTRICTID,DISTRICTNAME FROM DIC_REGION) AS E3 ON D.DISTRICT=E3.DISTRICTID
   LEFT OUTER JOIN (SELECT DISTINCT REGIONNAME,PROVINCEID FROM DIC_REGION) F ON SUBSTR(A.AREAKIND,2) = SUBSTR(F.PROVINCEID,2);

submit type: flink on yarn

与 DIC_REGION 表关联的部分显示为已完成。 我该如何处理以使检查点恢复正常？请。谢谢