某客戶在實施Nimber儲存雙活過程中，為了測試雙活是否生效，對儲存進行斷電測試。測試過程中，虛擬化未受到影響，虛擬機器都正常存活。但是安裝在兩臺Oracle Enterprise Linux 7.8作業系統上的 Oracle 11.2.0.4資料庫叢集，資料庫全部掛掉，雙活看起來是沒有達到效果。

在ASM的告警日誌中不斷提示can't access physical storage之類，

2022-03-19 15:32:26.315: [ OCRASM][3044980480]ASM Error Stack : ORA-27091: unable to queue I/O

ORA-15078: ASM diskgroup was forcibly dismounted

ORA-06512: at line 4

CRS的日誌中，出現如下資訊

2022-03-19 15:32:26.295: [ OCRRAW][3044980480]proprior: Header check from OCR device 0 offset 5857280 failed (26).

2022-03-19 15:32:26.295: [ OCRRAW][3044980480]proprior: Retrying buffer read from another mirror for disk group [+OCR] for block at offset [5857280]

2022-03-19 15:32:26.295: [ OCRASM][3044980480]proprasmres: Total 0 mirrors detected

2022-03-19 15:32:26.295: [ OCRASM][3044980480]proprasmres: Only 1 mirror found in this disk group.

2022-03-19 15:32:26.295: [ OCRASM][3044980480]proprasmres: Need to invoke checkdg. Mirror #0 has an invalid buffer.

2022-03-19 15:32:26.315: [ OCRASM][3044980480]proprasmres: kgfoControl returned error [8]

[ OCRASM][3044980480]SLOS : SLOS: cat=8, opn=kgfoCkDG01, dep=15032, loc=kgfokge

多路徑軟體設定與Nimber儲存推薦設定一致。最終諮詢了儲存原廠售後，建議修改ASM的例項引數_asm_hbeatiowait解決了CRS掛掉的問題。此隱含引數在11g中預設為15秒，在12cR2中為229秒

SQL> select ksppinm,ksppstvl,ksppdesc from x$ksppi x,x$ksppcv y where x.indx = y.indx and ksppinm='_asm_hbeatiowait';

KSPPINM KSPPSTVL KSPPDESC

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

_asm_hbeatiowait 229 number of secs to wait for PST Async Hbeat IO return

SQL> select * from v$version;

BANNER CON_ID

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

Oracle Database 12c Enterprise Edition Release 12.2.0.1.0 - 64bit Production 0

PL/SQL Release 12.2.0.1.0 - Production 0

CORE 12.2.0.1.0 Production 0

TNS for Linux: Version 12.2.0.1.0 - Production 0

NLSRTL Version 12.2.0.1.0 - Production 0

參考：

ORA-27091:unable to queue I/O &OCR

相關文章