苏大文正节点一 ORA-00603 ORA-27504 ORA-27300 ORA-27301 ORA-27302 BUG

 

Problem Description

---------------------------------------------------

Tue Sep 01 04:05:33 2020

skgxpvfynet: mtype: 61 process 417356 failed because of a resource problem in the OS. The OS has most likely run out of buffers (rval: 4)

Errors in file /u01/app/oracle/diag/rdbms/syntong/syntong1/trace/syntong1_w001_417356.trc (incident=96021):

ORA-00603: ORACLE server session terminated by fatal error

ORA-27504: IPC error creating OSD context

ORA-27300: OS system dependent operation:sendmsg failed with status: 105

ORA-27301: OS failure message: No buffer space available

ORA-27302: failure occurred at: sskgxpsnd2

Incident details in: /u01/app/oracle/diag/rdbms/syntong/syntong1/incident/incdir_96021/syntong1_w001_417356_i96021.trc

opidrv aborting process W001 ospid (417356) as a result of ORA-603

 

Error Codes

---------------------------------------------------

ORA-00603 ORA-27504 ORA-27300 ORA-27301 ORA-27302

 

根据MOS文档查询，发现匹配文档中的105错误：

 

Doc ID 579365.1:

 

https://support.oracle.com/epmos/faces/DocumentDisplay?_afrLoop=80914981328617&id=579365.1

 

STATUS 105 - ENOBUFS - No buffer space available

This error means that a socket cannot be created until resources are freed.

ORA-00603: ORACLE server session terminated by fatal error

ORA-27504: IPC error creating OSD context

ORA-27300: OS system dependent operation:sendmsg failed with status: 105

ORA-27301: OS failure message: No buffer space available

ORA-27302: failure occurred at: sskgxpsnd2

See: Oracle Linux: ORA-27301:OS Failure Message: No Buffer Space Available (Doc ID 2041723.1)

查看MTU：

MTU of loopback adapter is too high. We can use the command "netstat -in" see the current MTU size

Linux: #netstat -in

Kernel Interface table
   Iface   MTU   Met   RX-OK   RX-ERR RX-DRP RX-OVR TX-OK TX-ERR TX-DRP TX-OVR Flg
   eth0   1500    0    1371747  0      0      0     1858   0       0     0     BMRU
   lo     65536   0      46943  0      0      0    46943   0       0     0     LRU
   virbr0 1500    0          0  0      0      0       32   0       0     0     BMRU

文档提供的解决方案：

Doc ID 2041723.1:

https://support.oracle.com/epmos/faces/DocumentDisplay?_afrLoop=82221513985238&id=2041723.1

CAUSE

This happens due to less space available for network buffer reservation.

SOLUTION

1. On servers with High Physical Memory, the parameter vm.min_free_kbytes should be set in the order of 0.4% of total Physical Memory. This helps in keeping a larger range of defragmented memory pages available for network buffers reducing the probability of a low-buffer-space conditions.

*** For example, on a server which is having 256GB RAM, the parameter vm.min_free_kbytes should be set to 1073742 ***

On NUMA Enabled Systems, the value of vm.min_free_kbytes should be multiplied by the number of NUMA nodes since the value is to be split across all the nodes.

 

On NUMA Enabled Systems, the value of vm.min_free_kbytes = n * 0.4% of total Physical Memory. Here 'n' is the number of NUMA nodes.

2. Additionally, the MTU value should be modified as below

#ifconfig lo mtu 16436

To make the change persistent over reboot add the following line in the file /etc/sysconfig/network-scripts/ifcfg-lo :

MTU=16436

Save the file and restart the network service to load the changes

#service network restart

Note : While making the changes in CRS nodes, if network is restarted while CRS is up, it can hung CRS. So cluster services should be stopped prior to the network restart.

vm.min_free_kbytes

该参数表示Linux VM最低保留多少的空闲内存空间，当可用的内存低于配置参数时，系统会进行cache内存的回收，来进行内存的释放。

单位是kb

以下是我整理的苏大文正解决方案步骤：

SOLUTION:

 

 一、前期准备

1、检查集群及数据库运行情况

#su - grid

#crs_stat -t

#su - oracle

#sqlplus / as sysdba

SQL>select INST_ID,INSTANCE_NUMBER,INSTANCE_NAME,STATUS,DATABASE_STATUS,INSTANCE_ROLE from gv$Instance;

2、检查数据库备份情况

SQL>

col INPUT_BYTES_DISPLAY for a10

col OUTPUT_BYTES_DISPLAY for a10

col TIME_TAKEN_DISPLAY for a10

set line222

 

select input_type,

       status,

       to_char(start_time,

               'yyyy-mm-dd hh24:mi:ss'),

       to_char(end_time,

               'yyyy-mm-dd hh24:mi:ss'),

       input_bytes_display,

       output_bytes_display,

       time_taken_display

  from v$rman_backup_job_details

 where start_time > date '2020-09-10'

 order by 3 desc;

 

3、备份节点一/etc/sysctl.conf文件

#cp /etc/sysctl.conf /home/oracle/pst

 

二、配置调整

集群环境需要先关闭CRS集群服务，修改网络配置会Hang住CRS；修改系统参数需要重启数据库；

顺序关闭 数据库，集群 (共三套库：syntong1，carddb1，urpdb1）

1、关闭数据库：

#su - grid

#srvctl stop instance -d syntong -i syntong1

#srvctl stop instance -d syntong -i syntong2

#srvctl stop instance -d carddb -i carddb1

#srvctl stop instance -d carddb -i carddb2

#srvctl stop instance -d urpdb -i urpdb1

#srvctl stop instance -d urpdb -i urpdb2

#which crsctl

 

2、关闭集群：

#su - root

#cd /u01/app/grid/product/11.2.0/db_1/bin/crsctl

#crsctl stop cluster -all

 

节点一：（由于节点一报错ora-2700*，因此修改节点一）

3、修改网络配置：

修改配置文件

#vi /etc/sysconfig/network-scripts/ifcfg-lo

MTU=16436

重启网络服务

# systemctl restart network

 

4、修改系统参数：

设定 vm.min_free_kbytes 参数为物理内存的0.4%

本机内存大小为131357180 Kb，则配置参数大小为131357180*0.4%≈525429

#vi /etc/sysctl.conf

vm.min_free_kbytes = 525429

生效

#sysctl -p

顺序开启 集群，数据库 (共三套库：syntong1，carddb1，urpdb1）

5、开启集群：

#su - root

#cd /u01/app/grid/product/11.2.0/db_1/bin/crsctl

#crsctl start cluster -all

 

6、开启数据库：

#su - grid

#srvctl start instance -d syntong -i syntong1

#srvctl start instance -d syntong -i syntong2

#srvctl start instance -d carddb -i carddb1

#srvctl start instance -d carddb -i carddb2

#srvctl start instance -d urpdb -i urpdb1

#srvctl start instance -d urpdb -i urpdb2

检查集群及数据库运行情况

#su - grid

#crs_stat -t

#su - oracle

#sqlplus / as sysdba

SQL>select INST_ID,INSTANCE_NUMBER,INSTANCE_NAME,STATUS,DATABASE_STATUS,INSTANCE_ROLE from gv$Instance;

三、后续观察

查看是否继续出现ORA-2700*错误：

#su - oracle

#adrci

adrci>show problem

adrci>show incident

#cd $ORACLE_BASE/diag/rdbms/syntong/syntong1/trace

#tail -f alert_syntong1.log