[10] From 11 is switched to the
line 10, the management
cat / proc / drbd
normal state is:
[ODB-MlO the root @ ~] # cat / proc / drbd
Version: 8.3.2 (API: 88 / proto: 86- 90)
the GIT the hash-: Build dd7985327f146f33b86d4bff5ca8c94234ce840e by [email protected], 2011-08-31 02:05:57
0: CS: Connected RO: Secondary / Primary DS: to UpToDate / C to UpToDate R & lt ----
NS: 0 NR: DW 602 236 972: 602 236 972 DR: Al 0: 0 BM: 1026 LO: PE 0: 0 UA: 0 AP: 0 EP: WO. 1: OOS D: 0
- automatic switch script address
[waste] / etc /drbd.sh
--1 view
the clustat
--2 the clustat
order to check the data in the table
MySQL> the SELECT COUNT (*) from order_xxxx;
MySQL> the SELECT COUNT (*) from order_xxxx;
+ ---------- +
| COUNT (*) |
+ ---------- +
| 889 067 |
+ ---------- +
1 Row in the SET (0.34 sec)
to ensure that data can not be cut in the past lost
--3 10 executed 20.20.30.21 switch to the above command to switch
[]
to look at before the cut 1 which service is above
2 before switching to withdraw from all the mysql client window, otherwise it will affect the switching
clusvcadm -r mysqlservice -m odb-m10.xx.com
[Comment]
Start is clusvcadm -e webserver -m web1
Close clusvcadm -s <Service> -m <Node>
--4 view the status of
the clustat
--5 go from the library to check the data
mysql> select count (*) from order_xxxx;
- check 1
mysql -uadmin -p -P3307 -h20.20.30.20 --default-character-set=utf8
mysql -uadmin -p -P3307 -h127.0.0.1 --default-character-set=utf8
Check that truly landing up.
-- check 2
mk-table-checksum h=20.20.30.20,u=admin,p=MPxxxxxx,P=3307 h=20.20.30.23,u=admin,p=MPxxxxxx,P=3307 h=20.20.30.24,u=admin,p=MPxxxxxx,P=3307 h=20.20.30.25,u=admin,p=MPxxxxxx,P=3307 h=20.20.30.26,u=admin,p=MPxxxxxx,P=3307 h=20.20.30.22,u=admin,p=MPxxxxxx,P=3307
-d oshxx --count|mk-checksum-filter > /home/check/oshxx_2012-05-17.log