一套集群,一个节点启动正常,另外一个节点无法正常启动实例,启动异常节点alert日志

Tue Mar 07 19:07:29 2023
IPC Send timeout detected. Receiver ospid 6386 [
Tue Mar 07 19:07:29 2023
Errors in file /u01/app/oracle/diag/rdbms/xff/xff2/trace/xff2_lms0_6386.trc:
IPC Send timeout detected. Receiver ospid 6402 [
Tue Mar 07 19:07:29 2023
Errors in file /u01/app/oracle/diag/rdbms/xff/xff2/trace/xff2_lms4_6402.trc:
Tue Mar 07 19:07:29 2023
Received an instance abort message from instance 1
Please check instance 1 alert and LMON trace files for detail.
System state dump requested by (instance=2, osid=6384 (LMD0)), summary=[abnormal instance termination].
System State dumped to trace file /u01/app/oracle/diag/rdbms/xff/xff2/trace/xff2_diag_6374_20230307190729.trc
LMD0 (ospid: 6384): terminating the instance due to error 481
Dumping diagnostic data in directory=[cdmp_20230307190729],
      requested by (instance=2, osid=6384 (LMD0)), summary=[abnormal instance termination].
Instance terminated by LMD0, pid = 6384

正常节点alert日志

Tue Mar 07 19:02:07 2023
Reconfiguration started (old inc 20, new inc 22)
List of instances:
 1 2 (myinst: 1)
 Global Resource Directory frozen
 Communication channels reestablished
 Master broadcasted resource hash value bitmaps
 Non-local Process blocks cleaned out
Tue Mar 07 19:02:08 2023
 LMS 5: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
 LMS 2: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
 LMS 1: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
 LMS 0: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
 LMS 3: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
 LMS 7: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
Tue Mar 07 19:02:08 2023
Tue Mar 07 19:02:08 2023
 LMS 4: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
 LMS 6: 0 GCS shadows cancelled, 0 closed, 0 Xw survived
 Set master node info
 Submitted all remote-enqueue requests
 Dwn-cvts replayed, VALBLKs dubious
 All grantable enqueues granted
 Submitted all GCS remote-cache requests
 Fix write in gcs resources
Tue Mar 07 19:02:27 2023
IPC Send timeout detected. Sender: ospid 6936 [oracle@xffnode1.localdomain (PING)]
Receiver: inst 2 binc 441249706 ospid 59731
Tue Mar 07 19:07:29 2023
IPC Send timeout detected. Sender: ospid 6946 [oracle@xffnode1.localdomain (LMS0)]
Receiver: inst 2 binc 429479852 ospid 6386
Tue Mar 07 19:07:29 2023
IPC Send timeout detected. Sender: ospid 6962 [oracle@xffnode1.localdomain (LMS4)]
Receiver: inst 2 binc 429479854 ospid 6402
Tue Mar 07 19:07:29 2023
IPC Send timeout detected. Sender: ospid 6966 [oracle@xffnode1.localdomain (LMS5)]

通过上述日志,可以确认主要由于两个节点之间无法正常通讯,从而使得新节点无法加入到集群(无法完成集群重组),从而使得实例启动异常.一般出现这类情况最检查的就是私网异常,通过分析oswnetstat记录发现packet reassembles failed特别严重

---

一般出现该问题,考虑是由于ipfrag_*_thresh默认值不足导致,通过设置

net.ipv4.ipfrag_high_thresh = 16777216
net.ipv4.ipfrag_low_thresh = 15728640

临时请库成功,但是数据库实例重组时间依旧过长

---

packet reassembles failed依旧在增加,通过分析网卡情况发现网卡异常,采用haip(双万兆网卡)的其中一块网卡异常

---

为了数据库性能不收太大影响,临时禁用异常网卡,后续等网络层面解决之后再启用