Db2 troubleshooting tools
The db2pd command::
You can use the Db2 problem determination
tool, db2pd, to obtain troubleshooting information from database system memory
sets, without acquiring any lock or latch in the database. Since the db2pd tool
works directly with memory, it is possible to retrieve information that is
changing as it is being collected using minimal Db2 engine resources.
If the database is running in a Massively
Parallel Processing (MPP) setup, a pureScale configuration, or an HADR setup,
this command must be run on all the nodes or members of the cluster. However,
if you have logical partitions within one physical server in an MPP setup, it
collects information for all the logical nodes of the current host.
If you want to see the HADR database status,
you can execute the db2pd command either on the primary or on the standby
server.
Use the following command to execute the tool:
db2pd -db SAMPLE -hadr
Database Member 0 -- Database SAMPLE -- Active
-- Up 170 days 08:17:44 -- Date 2018-05-13-20.23.52.606937
HADR_ROLE = PRIMARY
REPLAY_TYPE = PHYSICAL
HADR_SYNCMODE = NEARSYNC
STANDBY_ID = 1
LOG_STREAM_ID = 0
HADR_STATE = PEER
HADR_FLAGS =
PRIMARY_MEMBER_HOST = DB2AXNODE001
PRIMARY_INSTANCE = db2inst1
PRIMARY_MEMBER = 0
STANDBY_MEMBER_HOST = DB2AXNODE002
STANDBY_INSTANCE = db2inst1
STANDBY_MEMBER = 0
HADR_CONNECT_STATUS = CONNECTED
HADR_CONNECT_STATUS_TIME = 11/24/2017
11:08:30.948572 (1511543310)
HEARTBEAT_INTERVAL(seconds) = 30
HEARTBEAT_MISSED = 3
HEARTBEAT_EXPECTED = 21499
HADR_TIMEOUT(seconds) = 120
TIME_SINCE_LAST_RECV(seconds) = 0
PEER_WAIT_LIMIT(seconds) = 300
LOG_HADR_WAIT_CUR(seconds) = 0.000
LOG_HADR_WAIT_RECENT_AVG(seconds) = 0.000681
LOG_HADR_WAIT_ACCUMULATED(seconds) =
787983.115
LOG_HADR_WAIT_COUNT = 529872503
SOCK_SEND_BUF_REQUESTED,ACTUAL(bytes) = 0,
262088
SOCK_RECV_BUF_REQUESTED,ACTUAL(bytes) = 0,
262088
PRIMARY_LOG_FILE,PAGE,POS = S0135197.LOG,
33536, 21171039888078
STANDBY_LOG_FILE,PAGE,POS = S0135197.LOG,
33521, 21171039826535
HADR_LOG_GAP(bytes) = 2508
STANDBY_REPLAY_LOG_FILE,PAGE,POS =
S0135197.LOG, 33520, 21171039822144
STANDBY_RECV_REPLAY_GAP(bytes) = 3728
PRIMARY_LOG_TIME = 05/13/2018 20:23:52.000000
(1526261032)
STANDBY_LOG_TIME = 05/13/2018 20:23:49.000000
(1526261029)
STANDBY_REPLAY_LOG_TIME = 05/13/2018
20:23:48.000000 (1526261028)
STANDBY_RECV_BUF_SIZE(pages) = 131072
STANDBY_RECV_BUF_PERCENT = 0
STANDBY_SPOOL_LIMIT(pages) = 33423360
STANDBY_SPOOL_PERCENT = 0
STANDBY_ERROR_TIME = NULL
PEER_WINDOW(seconds) = 300
PEER_WINDOW_END = 05/13/2018 20:28:50.000000
(1526261330)
READS_ON_STANDBY_ENABLED = N
If you want to extract all of the locking
information on the SAMPLE database, you can run the db2pd command without
actually locking or latching any resource in the database using the following
command:
db2pd -db SAMPLE -locks
Database Member 0 -- Database SAMPLE -- Active
-- Up 170 days 08:19:35 -- Date 2018-05-13-20.25.43.409995
Locks:
Address TranHdl Lockname Type Mode Sts Owner
Dur HoldCount Att ReleaseFlg rrIID
0x0A000300464B7B80 132
FFFA801500390006C0F6000252 RowLock ..X G 132 1 0 0x00200020 0x40000000 0
0x0A000300464CCA00 132
00CE00040000000143DD009952 RowLock ..X G 132 1 0 0x00200008 0x40000000 0
0x0A000300464E4800 132 00CE00040000000143DC00BA52
RowLock ..X G 132 1 0 0x00200008 0x40000000 0
0x0A000300464CDF80 132
00CE00040000000143DE002E52 RowLock ..X G 132 1 0 0x00200008 0x40000000 0
Comments
Post a Comment