Home >System Tutorial >LINUX >Linux does not restart to recognize newly mounted disks
The RAC database expands the storage space. The newly mounted disk is not recognized and cannot be seen through the fdisk -l command. After querying, you need to restart or scan to see the newly mounted disk. Today I will share the specific method.
I. Overview
The RAC database expands the storage space. The newly mounted disk is not recognized and cannot be seen through the fdisk -l command. After querying, you need to restart or scan to see the newly mounted disk. Today I will share the specific method.
Environment: OEL 6.4 and RAC 11.2.0.4
2. Key matters
Do not use /dev/sdX or /dev/dm-XX as the device name because this device name may change after a reboot
Identify device using WWID
Worldwide identifiers (WWID) can be used to reliably identify disk devices.
A WWID is a persistent, system-independent ID that the SCSI standard requires all SCSI devices to provide.
3. Display all available WWIDs on the system
Shell script to display WWID on your system #!/bin/bash # #Usage: As root user run : ./check_wwid.sh # for FILE in `find /dev -name "sd*" | sort` do WWID=`scsi_id --whitelisted --replace-whitespace --device=$FILE ` echo $FILE " WWID: " $WWID done $ sudo ./check_wwid.sh .. /dev/sdb WWID: 1ATA_VBOX_HARDDISK_VB81c4f844-94203a15 /dev/sdb1 WWID: 1ATA_VBOX_HARDDISK_VB81c4f844-94203a15 /dev/sdc WWID: 1ATA_VBOX_HARDDISK_VB2ed83c09-e66dbcfb /dev/sdc1 WWID: 1ATA_VBOX_HARDDISK_VB2ed83c09-e66dbcfb .. --> Of course WWIDs for partitions and for the complete disk are identical # cat 99-oracle-asmdevices.rules KERNEL=="sd*", SUBSYSTEM=="block", ENV{DEVTYPE}=="disk", ENV{ID_SERIAL}=="14f504e46494c45523264556151442d5261336d2d566e4250", OWNER="grid", GROUP="asmadmin", MODE="0660" # ls -ls /dev/sdr 0 brw-rw----. 1 grid asmadmin 65, 16 Mar 19 14:42 /dev/sdr Reload and Restart the udev rules # udevadm control --reload-rules # start_udev Starting udev: [ OK ] Verify disk protections: # ls -ltr /dev/asm* brw-rw----. 1 grid asmadmin 8, 17 Mar 15 10:03 /dev/asmdisk1_udev_sdb1 brw-rw----. 1 grid asmadmin 8, 33 Mar 15 10:03 /dev/asmdisk2_udev_sdc1 Redistribute 99-oracle-asmdevices.rules file to all cluster nodes and restat udev echo system on these nodes # scp 99-oracle-asmdevices.rules grac42:/etc/udev/rules.d # scp 99-oracle-asmdevices.rules grac43:/etc/udev/rules.d On each RAC node # udevadm control --reload-rules # start_udev # ls -ltr /dev/asm* brw-rw----. 1 grid asmadmin 8, 17 Mar 15 10:03 /dev/asmdisk1_udev_sdb1 brw-rw----. 1 grid asmadmin 8, 33 Mar 15 10:03 /dev/asmdisk2_udev_sdc1
4. UDev rules for disk partitions using /sbin/scsi_id (preferred mode because we can change the name)
Note this example mapping: /dev/sdb1 to /dev/asmdisk1_udev_sdb1:
# cat 99-oracle-asmdevices.rules KERNEL=="sd?1", BUS=="scsi", PROGRAM=="/sbin/scsi_id -g -u -d /dev/$parent", RESULT=="1ATA_VBOX_HARDDISK_VB81c4f844-94203a15", NAME= "asmdisk1_udev_sdb1", OWNER="grid", GROUP="asmadmin", MODE="0660" KERNEL=="sd?1", BUS=="scsi", PROGRAM=="/sbin/scsi_id -g -u -d /dev/$parent", RESULT=="1ATA_VBOX_HARDDISK_VB2ed83c09-e66dbcfb", NAME= "asmdisk1_udev_sdc1", OWNER="grid", GROUP="asmadmin", MODE="0660" # ls -l /dev/asmdisk1_udev_sdb1 /dev/asmdisk1_udev_sdc1 brw-rw----. 1 grid asmadmin 8, 17 Mar 19 13:53 /dev/asmdisk1_udev_sdb1 brw-rw----. 1 grid asmadmin 8, 33 Mar 19 13:53 /dev/asmdisk1_udev_sdc1
5. Disk group
(1) Expanding disk group
Find path:
select group_number,disk_number ,path from v$asm_disk;
Expanding disk group:
alter diskgroup data2 add disk '/dev/asm-diski'; ALTER DISKGROUP DATA2 REBALANCE POWER 10;
View is balanced completion:
select operation,est_minutes from v$asm_operation;
Modify it after completion:
alter diskgroup DATA rebalance power 1;
Or: do the balancing directly:
alter diskgroup DATA add disk '/dev/asm-diski' rebalance power 10;
The above is the detailed content of Linux does not restart to recognize newly mounted disks. For more information, please follow other related articles on the PHP Chinese website!