Detecting slow inserts
We have an application that does hundreds of inserts per second to a specific table.
Naturally we are worried about performance there.
Is there a way to detect when insert into the table takes over X seconds? Ideally the time this occured, the insert command and how long it took will be written to another table, so we can later analyze the results.
Note that we don't want to do extra inserts just for testing how long they will take, we want to monitor the performance of the real application inserts.
Is it possible at all? Does Oracle "know" how long the insert took?
One simple approach is to monitor wait events on relevant segments - your table and its indexes. So using Statspack you can identify (in a 15 minute period, for example) whether there are any buffer busy waits, row lock waits etc, and if so, how long they took. If you have the level right (7, I think) you can also see the cumulative cost of specific SQL including inserts (counts, plus elapsed time and cpu if TIMED_STATISTICS are switched on).
Statspack generally reports the top 5 or 10; you can easily modify that if you like, but if this insert isn't in the top 10, you're probably not interested in tuning it....
HTH
Regards Nigel
Similar Messages
-
We suddently started seeing some slow inserts on some test boxes. All boxes are on RedHat 5 and 10.2.0.1.
We have a multi-tier web application conencting to the DB using connection pool and inserting about a 100000 records in a table. They first do a select, if not found then its an insert. So to take the other pieces out of the puzzle. I wrote a small plssql, which does a select and then does an insert based on some random data generated.
When I ran this on my test system(db1) it ran about 37 sec, cpu usage was very low. When I ran this on the system that was reported to be slow(db2), it ran for about 30 minutes and CPU was pegged.
Then I think we let this system and when I tried today, on the system (db2), it completed in about 35 sec. This was wierd. So I re-created the schema (dropped the user) and then ran the test, it was slow again and the CPU was pegged.
So I thought well may be since the system was not touched (no inserts or any activity) the stats jobs ran, etc that could have been the reason. So ran stats on the database and the schema but no use, it was still slow.
So then I ran the PL/SQL block on another system (db3) where the schema was created but no iserts or and activty on this box for atleast a few days. But then it was extemely slow in this box as well and cpu was pegged.
All these boxes have the same table, indexes, they are all identical. Any reason why this could be happening and what else should I be looking for?
SET timing on;
DECLARE
uname NVARCHAR2 (60);
primid NVARCHAR2 (60);
num NUMBER;
BEGIN
FOR i IN 1 .. 100000
LOOP
uname := DBMS_RANDOM.STRING ('x', 20);
primid := DBMS_RANDOM.STRING ('x', 30);
DBMS_OUTPUT.put_line (uname || ' ==> ' || primid);
SELECT COUNT (*)
INTO num
FROM test
WHERE (deldate IS NULL)
AND id = 0
AND ( primid = primid
OR UPPER (username) = uname
OR uiname = uname
ORDER BY time DESC;
INSERT INTO test
VALUES (0, uname, uname, 1,
uname, primid
IF MOD (i, 200) = 0
THEN
COMMIT;
DBMS_OUTPUT.put_line ('Commited');
END IF;
END LOOP;
END;
/So I used
execute sys.dbms_support.start_trace_in_session(175, 9629, waits=>true, binds=>false);
Othere than a couple of these I don't see any other wait events, should I be looking for something else or am I doing the trace wrong. I think this is equivalent to what you suggested.
Elapsed times include waiting on following events:
Event waited on Times Max. Wait Total Waited
---------------------------------------- Waited ---------- ------------
log file sync 1 0.00 0.00
SQL*Net message to client 2 0.00 0.00
SQL*Net message from client 1 0.00 0.00
Elapsed times include waiting on following events:
Event waited on Times Max. Wait Total Waited
---------------------------------------- Waited ---------- ------------
log file switch completion 3 0.00 0.00 -
Slow inserts after log size increase
I have an app that does 10k-50k consecutive SELECTs and INSERTs as part of a loading operation. Recently, I enlarged the redo logs from 10M to 100M, which appears about right, based on the frequency of log switches (half a dozen/day). Trouble is, now the load takes 5 times longer (20-25 min vs. 5). Any thoughts? I thought redo logs that were too small caused poor performance, b/c of increased checkpoint activity, but I can't imagine why larger redo logs would slow inserts down.
RobLogically the only difference a larger online redo log should make is that the number of checkpoints invoked due to log switches should decrease. But a larger online redo log will take longer to archive.
Where did you place the larger online redo log and where does the archive redo log go to? Is this performance decrease consistent or does it appear when certain redo logs are in use?
Even if your logical disk are different it could be the physical disk for some of your files are now the same where before this was not true.
The suggested statspack is probably a good place to start but you might also want to take some OS IO stats during the run also.
HTH -- Mark D Powell -- -
hi,
my below insert scripts takes lot of time , even dbcontrol displays high hash value for this insert
insert into id_air_e (aire_key,aire_0) values (:b0,:b1)
additional details:
Name ID_AIR_E
Tablespace USERS
Organization Standard, Heap Organized
SQL> desc id_air_e
Name Null? Type
AIRE_KEY NOT NULL NUMBER
AIRE_0 VARCHAR2(240)
Constraints
Name Type Table Columns Disabled Deferrable Initially Deferred Validate RELY
AIR_E_FKEY FOREIGN AIRE_KEY NO NO NO YES NO
SYS_C004747 CHECK AIRE_KEY NO NO NO YES NO
Indexes
Number of Indexes 0
sql> select count(1) from id_air_e
201456
any suggestions to increase the insert timeHi,
i created index on FK, dba_blockers view has no records, SGA is also well sized.
still i face v slow insert in this table
Name ID_AIR_E
Organization Standard, Heap Organized
Columns:
AIRE_KEY NUMBER
AIRE_0 VARCHAR2 240
Constraints:
AIR_E_FKEY FOREIGN AIRE_KEY NO NO NO YES NO
SYS_C004747 CHECK AIRE_KEY NO NO NO YES NO
Indexes:
AE001T3.ID_AIR_E_INDX on AIRE_KEY
SGA Allocation (MB)
Shared Pool 304
Buffer Cache 1664
Large Pool 16
Java Pool 16
Other 48
PGA : 1800 MB -
Jdbc thin driver bulk binding slow insertion performance problem
Hello All,
We have a third party application reporting slow insertion performance, while I traced the session and found out most of elapsed time for one insert execution is sql*net more data from client, it appears bulk binding is being used here because one execution has 200 rows inserted. I am wondering whether this has something to do with their jdbc thin driver(10.1.0.2 version) and our database version 9205. Do you have any similar experience on this, what other possible directions should I explore?
here is the trace report from 10046 event, I hide table name for privacy reason.
Besides, I tested bulk binding in PL/SQL to insert 200 rows in one execution, no problem at all. Network folks confirm that network should not be an issue as well, ping time from app server to db server is sub milisecond and they are in the same data center.
INSERT INTO ...
values
(:1, :2, :3, :4, :5, :6, :7, :8, :9, :10, :11, :12, :13, :14, :15, :16, :17,
:18, :19, :20, :21, :22, :23, :24, :25, :26, :27, :28, :29, :30, :31, :32,
:33, :34, :35, :36, :37, :38, :39, :40, :41, :42, :43, :44, :45)
call count cpu elapsed disk query current rows
Parse 1 0.00 0.00 0 0 0 0
Execute 1 0.02 14.29 1 94 2565 200
Fetch 0 0.00 0.00 0 0 0 0
total 2 0.02 14.29 1 94 2565 200
Misses in library cache during parse: 1
Optimizer goal: CHOOSE
Parsing user id: 25
Elapsed times include waiting on following events:
Event waited on Times Max. Wait Total Waited
---------------------------------------- Waited ---------- ------------
SQL*Net more data from client 28 6.38 14.19
db file sequential read 1 0.02 0.02
SQL*Net message to client 1 0.00 0.00
SQL*Net message from client 1 0.00 0.00
********************************************************************************I have exactly the same problem, I tried to find out what is going on, changed several JDBC Drivers on AIX, but no hope, I also have ran the process on my laptop which produced a better and faster performance.
Therefore I made a special solution ( not practical) by creating flat files and defining the data as an external table, the oracle will read the data in those files as they were data inside a table, this gave me very fast insertion into the database, but still I am looking for an answer for your question here. Using Oracle on AIX machine is a normal business process followed by a lot of companies and there must be a solution for this. -
Jdbc thin driver and bulk binding slow insertion performance
Hello All,
We have a third party application reporting slow insertion performance, while I traced the session and found out most of elapsed time for one insert execution is sql*net more data from client, it appears bulk binding is being used here because one execution has 200 rows inserted. I am wondering whether this has something to do with their jdbc thin driver(10.1.0.2 version) and our database version 9205. Do you have any similar experience on this, what other possible directions should I explore?
here is the trace report from 10046 event, I hide table name for privacy reason.
Besides, I tested bulk binding in PL/SQL to insert 200 rows in one execution, no problem at all. Network folks confirm that network should not be an issue as well, ping time from app server to db server is sub milisecond and they are in the same data center.
INSERT INTO ...
values
(:1, :2, :3, :4, :5, :6, :7, :8, :9, :10, :11, :12, :13, :14, :15, :16, :17,
:18, :19, :20, :21, :22, :23, :24, :25, :26, :27, :28, :29, :30, :31, :32,
:33, :34, :35, :36, :37, :38, :39, :40, :41, :42, :43, :44, :45)
call count cpu elapsed disk query current rows
Parse 1 0.00 0.00 0 0 0 0
Execute 1 0.02 14.29 1 94 2565 200
Fetch 0 0.00 0.00 0 0 0 0
total 2 0.02 14.29 1 94 2565 200
Misses in library cache during parse: 1
Optimizer goal: CHOOSE
Parsing user id: 25
Elapsed times include waiting on following events:
Event waited on Times Max. Wait Total Waited
---------------------------------------- Waited ---------- ------------
SQL*Net more data from client 28 6.38 14.19
db file sequential read 1 0.02 0.02
SQL*Net message to client 1 0.00 0.00
SQL*Net message from client 1 0.00 0.00
********************************************************************************I have exactly the same problem, I tried to find out what is going on, changed several JDBC Drivers on AIX, but no hope, I also have ran the process on my laptop which produced a better and faster performance.
Therefore I made a special solution ( not practical) by creating flat files and defining the data as an external table, the oracle will read the data in those files as they were data inside a table, this gave me very fast insertion into the database, but still I am looking for an answer for your question here. Using Oracle on AIX machine is a normal business process followed by a lot of companies and there must be a solution for this. -
Extremely slow inserts/updates reported after HW upgrade...
Hi gyus,
I'll try to be as descriptive as I can. It's this project in which we have to move circa 6 mission critical (24x7) and mostly OLTP databases from MS Windows 2003 (DB on local disks) to HP-UX IA (CA metrocluster, HP XP 12000 disk array) - all ORA10gR2 10.2.0.4. And everything was perfect until we moved this XYZ database...
Almost immediately users reported "considerable" performance degradation. According to 3rd party application log they get almost 40 secs. instead of previously recorded 10.
We, I mean Oracle and HP specialists, haven't noticed/recorded any significant peeks/bottlenecks (RAM, CPU, Disk I/O).
Feel free to check 3 AWR reports and the init.ora at [http://www.mediafire.com/?sharekey=0269c9bc606747b47f7ec40ada4772a6e04e75f6e8ebb871]
1_awrrpt_standard.txt - standard workload during 8 hours (peek hours are from 8-12AM)
2_awrrpt_2hrs_ca.txt - standard workload during 2 peek hours (8-10)
3_awrrpt_2hrs_noca.txt - standard workload during 2 peek hours (10-12) with CA disk mirroring disabled
Of course, I've checked the ADDM reports - and first, I'd like to ask why ADDM keeps on reporting the following (on all database instances on this
cluster node):
FINDING 1: 100% impact (310 seconds)
Significant virtual memory paging was detected on the host operating system.
RECOMMENDATION 1: Host Configuration, 100% benefit (310 seconds)
Is it just some kind of false alarm (like we use to get on MS Windows)? Both nodes are running on 32gigs of RAM
with roughly more than 10gigs constantly free.
Second, as ADDM reported:
FINDING 2: 44% impact (135 seconds)
Waits on event "log file sync" while performing COMMIT and ROLLBACK operations
were consuming significant database time.
we've tried to split CA disk mirroring, using RAID 10 for redo log file disks etc. etc. No substantial performance gain was reported from users (though I've noticed some in AWR reports).
Despite confusing app. users' feedback I'm nearly sure that our bottleneck are redo log file disks. Why? Previously (old HW) we had 1-3 ms avg wait on log file sync and log file parallel write and now (new HW, RAID5/RAID10 we've tested both) - it's 8 ms or even more. We were able to get 2ms only with CA switched off (HP etrocluster
disk array mirroring).
And that brings up two new questions:
1. Does redo log group mirroring (2 on 2 separate disks vs. 1 on 1 disk) have any
significant impact on abovementioned wait events? I mean what performance gain
could I expect when I drop all "secondary" redo log members?
2. Why do we get almost identical response times when we run bulk insert/update tests (say
1000000 rows) against old and new DB/HW?
Thanks in advance,
tOPsEEK
Edited by: smutny on 1.11.2009 17:39
Edited by: smutny on 1.11.2009 17:46Hi again,
so here's the actual AWR report (1 minute window while the most problematic operation took place). I think it's becoming clear we have to deal with slow redo log writes...
WORKLOAD REPOSITORY report for
DB Name DB Id Instance Inst Num Release RAC Host
... 294736156 ... 1 10.2.0.4.0 NO ...
Snap Id Snap Time Sessions Curs/Sess
Begin Snap: 1254 02-Nov-09 10:07:45 91 46.4
End Snap: 1255 02-Nov-09 10:08:47 91 46.4
Elapsed: 1.04 (mins)
DB Time: 0.51 (mins)
Cache Sizes
~~~~~~~~~~~ Begin End
Buffer Cache: 576M 576M Std Block Size: 8K
Shared Pool Size: 912M 912M Log Buffer: 14,372K
Load Profile
~~~~~~~~~~~~ Per Second Per Transaction
Redo size: 163,781.94 4,575.45
Logical reads: 1,677.15 46.85
Block changes: 1,276.32 35.66
Physical reads: 1.99 0.06
Physical writes: 1.16 0.03
User calls: 426.41 11.91
Parses: 20.74 0.58
Hard parses: 0.19 0.01
Sorts: 2.38 0.07
Logons: 0.00 0.00
Executes: 386.76 10.80
Transactions: 35.80
% Blocks changed per Read: 76.10 Recursive Call %: 31.51
Rollback per transaction %: 0.00 Rows per Sort: 369.98
Instance Efficiency Percentages (Target 100%)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Buffer Nowait %: 100.00 Redo NoWait %: 100.00
Buffer Hit %: 99.88 In-memory Sort %: 100.00
Library Hit %: 99.90 Soft Parse %: 99.07
Execute to Parse %: 94.64 Latch Hit %: 100.00
Parse CPU to Parse Elapsd %: 112.50 % Non-Parse CPU: 99.11
Shared Pool Statistics Begin End
Memory Usage %: 88.90 88.87
% SQL with executions>1: 98.74 99.39
% Memory for SQL w/exec>1: 95.35 97.75
Top 5 Timed Events Avg %Total
~~~~~~~~~~~~~~~~~~ wait Call
Event Waits Time (s) (ms) Time Wait Class
log file parallel write 2,228 21 10 69.4 System I/O
log file sync 2,220 21 10 69.2 Commit
CPU time 20 65.5
SQL*Net break/reset to client 2,106 1 1 3.4 Applicatio
db file sequential read 131 0 4 1.5 User I/O
Time Model Statistics DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
-> Total time in database user-calls (DB Time): 30.9s
-> Statistics including the word "background" measure background process
time, and so do not contribute to the DB time statistic
-> Ordered by % or DB time desc, Statistic name
Statistic Name Time (s) % of DB Time
DB CPU 20.2 65.5
sql execute elapsed time 9.1 29.4
PL/SQL execution elapsed time 0.3 1.0
parse time elapsed 0.2 .5
hard parse elapsed time 0.1 .5
repeated bind elapsed time 0.0 .0
DB time 30.9 N/A
background elapsed time 22.2 N/A
background cpu time 0.4 N/A
Wait Class DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
-> s - second
-> cs - centisecond - 100th of a second
-> ms - millisecond - 1000th of a second
-> us - microsecond - 1000000th of a second
-> ordered by wait time desc, waits desc
Avg
%Time Total Wait wait Waits
Wait Class Waits -outs Time (s) (ms) /txn
System I/O 3,213 .0 22 7 1.4
Commit 2,220 .0 21 10 1.0
Application 2,106 .0 1 1 0.9
User I/O 134 .0 0 4 0.1
Network 29,919 .0 0 0 13.4
Wait Events DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
-> s - second
-> cs - centisecond - 100th of a second
-> ms - millisecond - 1000th of a second
-> us - microsecond - 1000000th of a second
-> ordered by wait time desc, waits desc (idle events last)
Avg
%Time Total Wait wait Waits
Event Waits -outs Time (s) (ms) /txn
log file parallel write 2,228 .0 21 10 1.0
log file sync 2,220 .0 21 10 1.0
SQL*Net break/reset to clien 2,106 .0 1 1 0.9
db file sequential read 131 .0 0 4 0.1
control file parallel write 44 .0 0 9 0.0
db file parallel write 44 .0 0 4 0.0
SQL*Net more data to client 3,397 .0 0 0 1.5
SQL*Net message to client 26,509 .0 0 0 11.9
control file sequential read 897 .0 0 0 0.4
SQL*Net more data from clien 13 .0 0 0 0.0
direct path write 3 .0 0 0 0.0
SQL*Net message from client 26,510 .0 1,493 56 11.9
Streams AQ: qmn slave idle w 2 .0 55 27412 0.0
Streams AQ: qmn coordinator 4 50.0 55 13706 0.0
PL/SQL lock timer 10 100.0 49 4897 0.0
Background Wait Events DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
-> ordered by wait time desc, waits desc (idle events last)
Avg
%Time Total Wait wait Waits
Event Waits -outs Time (s) (ms) /txn
log file parallel write 2,228 .0 21 10 1.0
control file parallel write 44 .0 0 9 0.0
db file parallel write 44 .0 0 4 0.0
control file sequential read 71 .0 0 0 0.0
rdbms ipc message 2,412 7.7 525 218 1.1
pmon timer 20 100.0 59 2929 0.0
Streams AQ: qmn slave idle w 2 .0 55 27412 0.0
Streams AQ: qmn coordinator 4 50.0 55 13706 0.0
Operating System Statistics DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
Statistic Total
AVG_BUSY_TIME 847
AVG_IDLE_TIME 5,362
AVG_IOWAIT_TIME 2,692
AVG_SYS_TIME 295
AVG_USER_TIME 549
BUSY_TIME 3,396
IDLE_TIME 21,457
IOWAIT_TIME 10,776
SYS_TIME 1,190
USER_TIME 2,206
LOAD 0
OS_CPU_WAIT_TIME 192,401,000
RSRC_MGR_CPU_WAIT_TIME 0
VM_IN_BYTES 40,960
VM_OUT_BYTES 335,872
PHYSICAL_MEMORY_BYTES 34,328,276,992
NUM_CPUS 4
NUM_CPU_SOCKETS 4
Instance Activity Stats DB/Inst: ROVEPP/rovepp Snaps: 1254-1255
Statistic Total per Second per Trans
CPU used by this session 984 15.8 0.4
CPU used when call started 344 5.5 0.2
CR blocks created 4 0.1 0.0
Cached Commit SCN referenced 208 3.3 0.1
Commit SCN cached 0 0.0 0.0
DB time 2,589 41.6 1.2
DBWR checkpoint buffers written 69 1.1 0.0
DBWR checkpoints 0 0.0 0.0
DBWR object drop buffers written 0 0.0 0.0
DBWR tablespace checkpoint buffe 0 0.0 0.0
DBWR thread checkpoint buffers w 0 0.0 0.0
DBWR transaction table writes 8 0.1 0.0
DBWR undo block writes 15 0.2 0.0
IMU CR rollbacks 0 0.0 0.0
IMU Flushes 1,156 18.6 0.5
IMU Redo allocation size 996,048 16,017.2 447.5
IMU commits 2,100 33.8 0.9
IMU contention 0 0.0 0.0
IMU ktichg flush 0 0.0 0.0
IMU pool not allocated 0 0.0 0.0
IMU recursive-transaction flush 0 0.0 0.0
IMU undo allocation size 22,402,560 360,250.9 10,064.0
IMU- failed to get a private str 0 0.0 0.0
Misses for writing mapping 0 0.0 0.0
SQL*Net roundtrips to/from clien 26,480 425.8 11.9
active txn count during cleanout 34 0.6 0.0
application wait time 106 1.7 0.1
background checkpoints completed 0 0.0 0.0
background checkpoints started 0 0.0 0.0
background timeouts 199 3.2 0.1
branch node splits 0 0.0 0.0
buffer is not pinned count 13,919 223.8 6.3
buffer is pinned count 19,483 313.3 8.8
bytes received via SQL*Net from 884,016 14,215.7 397.1
bytes sent via SQL*Net to client 9,602,642 154,418.1 4,313.9
calls to get snapshot scn: kcmgs 13,641 219.4 6.1
calls to kcmgas 3,029 48.7 1.4
calls to kcmgcs 56 0.9 0.0
change write time 8 0.1 0.0
cleanout - number of ktugct call 42 0.7 0.0
cleanouts and rollbacks - consis 0 0.0 0.0
cleanouts only - consistent read 0 0.0 0.0
cluster key scan block gets 1,100 17.7 0.5
cluster key scans 1,077 17.3 0.5
commit batch/immediate performed 1 0.0 0.0
commit batch/immediate requested 1 0.0 0.0
commit cleanout failures: block 0 0.0 0.0
commit cleanout failures: buffer 0 0.0 0.0
commit cleanout failures: callba 4 0.1 0.0
commit cleanout failures: cannot 0 0.0 0.0
commit cleanouts 9,539 153.4 4.3
commit cleanouts successfully co 9,535 153.3 4.3
commit immediate performed 1 0.0 0.0
commit immediate requested 1 0.0 0.0
commit txn count during cleanout 26 0.4 0.0
concurrency wait time 0 0.0 0.0
consistent changes 264 4.3 0.1
consistent gets 48,659 782.5 21.9
consistent gets - examination 26,952 433.4 12.1
consistent gets direct 2 0.0 0.0
consistent gets from cache 48,657 782.4 21.9
cursor authentications 2 0.0 0.0
data blocks consistent reads - u 4 0.1 0.0
db block changes 79,369 1,276.3 35.7
db block gets 55,636 894.7 25.0
db block gets direct 3 0.1 0.0
db block gets from cache 55,633 894.6 25.0
deferred (CURRENT) block cleanou 4,768 76.7 2.1
dirty buffers inspected 0 0.0 0.0
enqueue conversions 15 0.2 0.0
enqueue releases 9,967 160.3 4.5
enqueue requests 9,967 160.3 4.5
enqueue timeouts 0 0.0 0.0
enqueue waits 0 0.0 0.0
execute count 24,051 386.8 10.8
failed probes on index block rec 0 0.0 0.0
frame signature mismatch 0 0.0 0.0
free buffer inspected 680 10.9 0.3
free buffer requested 1,297 20.9 0.6
heap block compress 11 0.2 0.0
hot buffers moved to head of LRU 1,797 28.9 0.8
immediate (CR) block cleanout ap 0 0.0 0.0
immediate (CURRENT) block cleano 2,274 36.6 1.0
index crx upgrade (positioned) 47 0.8 0.0
index fast full scans (full) 0 0.0 0.0
index fetch by key 10,326 166.1 4.6
index scans kdiixs1 6,071 97.6 2.7
leaf node 90-10 splits 14 0.2 0.0
leaf node splits 18 0.3 0.0
lob reads 0 0.0 0.0
lob writes 198 3.2 0.1
lob writes unaligned 176 2.8 0.1
logons cumulative 0 0.0 0.0
messages received 2,272 36.5 1.0
messages sent 2,272 36.5 1.0
no buffer to keep pinned count 0 0.0 0.0
no work - consistent read gets 21,083 339.0 9.5
opened cursors cumulative 1,290 20.7 0.6
parse count (failures) 0 0.0 0.0
parse count (hard) 12 0.2 0.0
parse count (total) 1,290 20.7 0.6
parse time cpu 18 0.3 0.0
parse time elapsed 16 0.3 0.0
physical read IO requests 124 2.0 0.1
physical read bytes 1,015,808 16,335.0 456.3
physical read total IO requests 1,030 16.6 0.5
physical read total bytes 15,785,984 253,851.1 7,091.6
physical read total multi block 0 0.0 0.0
physical reads 124 2.0 0.1
physical reads cache 122 2.0 0.1
physical reads cache prefetch 0 0.0 0.0
physical reads direct 2 0.0 0.0
physical reads direct (lob) 0 0.0 0.0
physical reads direct temporary 0 0.0 0.0
physical reads prefetch warmup 0 0.0 0.0
physical write IO requests 47 0.8 0.0
physical write bytes 589,824 9,484.8 265.0
physical write total IO requests 4,591 73.8 2.1
physical write total bytes 25,374,720 408,045.5 11,399.3
physical write total multi block 4,461 71.7 2.0
physical writes 72 1.2 0.0
physical writes direct 3 0.1 0.0
physical writes direct (lob) 3 0.1 0.0
physical writes direct temporary 0 0.0 0.0
physical writes from cache 69 1.1 0.0
physical writes non checkpoint 18 0.3 0.0
pinned buffers inspected 1 0.0 0.0
prefetch warmup blocks aged out 0 0.0 0.0
prefetched blocks aged out befor 0 0.0 0.0
process last non-idle time 0 0.0 0.0
recursive calls 12,197 196.1 5.5
recursive cpu usage 747 12.0 0.3
redo blocks written 11,398 183.3 5.1
redo buffer allocation retries 0 0.0 0.0
redo entries 6,920 111.3 3.1
redo log space requests 0 0.0 0.0
redo log space wait time 0 0.0 0.0
redo ordering marks 96 1.5 0.0
redo size 10,184,944 163,781.9 4,575.5
redo subscn max counts 811 13.0 0.4
redo synch time 2,190 35.2 1.0
redo synch writes 2,220 35.7 1.0
redo wastage 1,377,920 22,158.0 619.0
redo write time 2,192 35.3 1.0
redo writer latching time 0 0.0 0.0
redo writes 2,228 35.8 1.0
rollback changes - undo records 24 0.4 0.0
rollbacks only - consistent read 4 0.1 0.0
rows fetched via callback 1,648 26.5 0.7
session connect time 0 0.0 0.0
session cursor cache hits 1,242 20.0 0.6
session logical reads 104,295 1,677.2 46.9
session pga memory 2,555,904 41,101.0 1,148.2
session pga memory max 0 0.0 0.0
session uga memory 123,488 1,985.8 55.5
session uga memory max 0 0.0 0.0
shared hash latch upgrades - no 66 1.1 0.0
sorts (disk) 0 0.0 0.0
sorts (memory) 148 2.4 0.1
sorts (rows) 54,757 880.5 24.6
sql area evicted 86 1.4 0.0
sql area purged 0 0.0 0.0
summed dirty queue length 0 0.0 0.0
switch current to new buffer 596 9.6 0.3
table fetch by rowid 9,173 147.5 4.1
table fetch continued row 0 0.0 0.0
table scan blocks gotten 982 15.8 0.4
table scan rows gotten 154,079 2,477.7 69.2
table scans (cache partitions) 0 0.0 0.0
table scans (long tables) 0 0.0 0.0
table scans (short tables) 59 1.0 0.0
total number of times SMON poste 0 0.0 0.0
transaction rollbacks 1 0.0 0.0
undo change vector size 3,990,136 64,164.5 1,792.5
user I/O wait time 49 0.8 0.0
user calls 26,517 426.4 11.9
user commits 2,226 35.8 1.0
user rollbacks 0 0.0 0.0
workarea executions - onepass 0 0.0 0.0
workarea executions - optimal 204 3.3 0.1
write clones created in backgrou 0 0.0 0.0
write clones created in foregrou 0 0.0 0.0
-------------------------------------------------------------... and what's even more interesting is the report that I've got using this great Tanel Poder's session snapper script. Take a look at these numbers (excerpt):
SID USERNAME TYPE STATISTIC DELTA HDELTA/SEC %TIME
668 ROVE_EDA WAIT log file sync 2253426 450.69ms 45.1%
668 ROVE_EDA WAIT log file sync 2140618 428.12ms 42.8%
668 ROVE_EDA WAIT log file sync 2088327 417.67ms 41.8%
668 ROVE_EDA WAIT log file sync 2184408 364.07ms 36.4%
668 ROVE_EDA WAIT log file sync 2117470 352.91ms 35.3%
668 ROVE_EDA WAIT log file sync 2051280 341.88ms 34.2%
668 ROVE_EDA WAIT log file sync 1595019 265.84ms 26.6%
668 ROVE_EDA WAIT log file sync 612034 122.41ms 12.2%
668 ROVE_EDA WAIT log file sync 2162980 432.6ms 43.3%
668 ROVE_EDA WAIT log file sync 2071811 345.3ms 34.5%
668 ROVE_EDA WAIT log file sync 2004571 334.1ms 33.4%
668 ROVE_EDA WAIT db file sequential read 28401 5.68ms .6%
668 ROVE_EDA WAIT db file sequential read 29028 4.84ms .5%
668 ROVE_EDA WAIT db file sequential read 24846 4.14ms .4%
668 ROVE_EDA WAIT db file sequential read 24323 4.05ms .4%
668 ROVE_EDA WAIT db file sequential read 17026 3.41ms .3%
668 ROVE_EDA WAIT db file sequential read 6736 1.35ms .1%
668 ROVE_EDA WAIT db file sequential read 33028 5.5ms .6%
764 (LGWR) WAIT log file parallel write 2236748 447.35ms 44.7%
764 (LGWR) WAIT log file parallel write 2150825 430.17ms 43.0%
764 (LGWR) WAIT log file parallel write 2139532 427.91ms 42.8%
764 (LGWR) WAIT log file parallel write 2119086 423.82ms 42.4%
764 (LGWR) WAIT log file parallel write 2134938 355.82ms 35.6%
764 (LGWR) WAIT log file parallel write 2083649 347.27ms 34.7%
764 (LGWR) WAIT log file parallel write 2034998 339.17ms 33.9%
764 (LGWR) WAIT log file parallel write 1996050 332.68ms 33.3%
764 (LGWR) WAIT log file parallel write 1797057 299.51ms 30.0%
764 (LGWR) WAIT log file parallel write 555403 111.08ms 11.1%
764 (LGWR) WAIT log file parallel write 277875 46.31ms 4.6%
764 (LGWR) WAIT log file parallel write 2067591 344.6ms 34.5%Where SID=668 is the session we've been looking for... OK, we've got to get back to monitoring disk array and the corresponding network components.
tOPsEEK -
Using NI-VISA to detect USB insertion/removal
Is there a way to use NI-VISA to detect when a USB device has been inserted or removed?
I am currently using Windows API in my LabWindows application to capture the WM_DEVICECHANGE event. When this event is called, I check to see if the USB devices are still connected to my PC by using viFindRsrc function. On Windows XP 32-bit, this approached worked successfully. On my Windows 7 64-bit system this is no longer working. After I remove the USB device, the viFindRsrc function is still indicating that the device is connected even though it isn't.
Also, when a USB device is controlled by a NI-VISA driver, I cannot capture the Windows DBT_DEVICEARRIVAL and DBT_DEVICEREMOVECOMPLETE events. I am able to see these events for other USB devices that are not controlled by a NI-VISA driver.
Solved!
Go to Solution.I solved the issue.
The behavior of the NI-VISA driver requires that you register to receive Windows notifications. The other USB devices that weren't controlled by a NI-VISA driver didn't require me to register to receive the DBT_DEVICEARRIVAL and DBT_DEVICEREMOVECOMPLETE events. I received them by default...
Here is the code to register USB events
Here is the code where I register the USB events:
DEV_BROADCAST_DEVICEINTERFACE NotificationFilter;
intptr_t postHandle = 0;
/* Windows HWND */
HWND hWnd = 0;
long long unsigned int hWndVal = 0;
/* Get HWND of LabWindows\CVI panel */
GetPanelAttribute(panelHandle,ATTR_SYSTEM_WINDOW_HANDLE,&hWndVal);
hWnd = (HWND)hWndVal;
* USB Raw Device
* Device Interface Class GUID = {a5dcbf10-6530-11d2-901f-00c04fb951ed}
GUID_CLASS_USB_DEVICE.Data1 = 0xA5DCBF10L;
GUID_CLASS_USB_DEVICE.Data2 = 0x6530;
GUID_CLASS_USB_DEVICE.Data3 = 0x11D2;
GUID_CLASS_USB_DEVICE.Data4[0] = 0x90;
GUID_CLASS_USB_DEVICE.Data4[1] = 0x1F;
GUID_CLASS_USB_DEVICE.Data4[2] = 0x00;
GUID_CLASS_USB_DEVICE.Data4[3] = 0xC0;
GUID_CLASS_USB_DEVICE.Data4[4] = 0x4F;
GUID_CLASS_USB_DEVICE.Data4[5] = 0xB9;
GUID_CLASS_USB_DEVICE.Data4[6] = 0x51;
GUID_CLASS_USB_DEVICE.Data4[7] = 0xED;
NotificationFilter.dbcc_size = sizeof(DEV_BROADCAST_DEVICEINTERFACE);
NotificationFilter.dbcc_devicetype = DBT_DEVTYP_DEVICEINTERFACE;
NotificationFilter.dbcc_classguid = GUID_CLASS_USB_DEVICE;
* Installs a callback function for WM_DEVICECHANGE Windows message posted
* or sent to a LabWindows/CVI panel.
InstallWinMsgCallback (panelHandle, WM_DEVICECHANGE,
WindowsDeviceDetection, VAL_MODE_INTERCEPT, NULL, &postHandle);
* Register device notification for USB Raw Device interface class GUID
RegisterDeviceNotification(hWnd, &NotificationFilter,
DEVICE_NOTIFY_WINDOW_HANDLE);
Here is the code that responds:
static int CVICALLBACK WindowsDeviceDetection (int panelHandle, int message,
unsigned int* wParam, unsigned int* lParam, void* callbackData)
PDEV_BROADCAST_HDR lpdb = (PDEV_BROADCAST_HDR)lParam;
switch (message)
/* Received Windows WM_DEVICECHANGE message */
case WM_DEVICECHANGE:
switch (*wParam)
* Received Windows device-change event "DBT_DEVNODES_CHANGED".
* Device has been added or removed from the system.
case DBT_DEVNODES_CHANGED:
printf("[%s] DBT_DEVNODES_CHANGED\n",TimeStr());
break;
* Received Windows device-change event "DBT_DEVICEARRIVAL".
case DBT_DEVICEARRIVAL:
printf("[%s] DBT_DEVICEARRIVAL\n",TimeStr());
break;
* Received Windows device-change event "DBT_DEVICEREMOVECOMPLETE".
case DBT_DEVICEREMOVECOMPLETE:
printf("[%s] DBT_DEVICEREMOVECOMPLETE\n",TimeStr());
break;
default:
break;
return 0; -
Dear Professionals:
The insert and query become very slow and somtimes hang. It wasn't like this before .. Network people added 100 pc to the network .. most of them not using the databse only internet and we are all in the same network .. Can this slow down the database Oracle 9.2.0.1.0 OS w2k server ? And how to know that the slowness is from network and not from query or inserts ?
Ahmed.Hi,
>>Network people added 100 pc to the network .. most of them not using the databse only internet. Can this slow down the database Oracle 9.2.0.1.0 OS w2k server ?
Maybe yes, maybe not, maybe a network performance problem ...
If you try to execute these DML's (insert, ...) and querys directly on the Server, what's happen ?
Cheers -
Slow Inserts Using IKM SQL to SQL Append
Hi all.
I'm in a big, big trouble.
I have a interface that gets that from the source and insert in the target (obvious)... but the unique way to insert into target is using SQL to SQL Append.
Unfortunately, the inserts are about 10 rows/sec. Very, very, very slow.
There are any ways to do the Agent faster than this?
Regards.
Luiz Araujo (a.k.a. Buchu_VP)My Staging Area is different from the target area, but unfortunately I'm not able to change this 'cause my sources are accessed via DBLink and there are restrictions to it, enabling only the IP and HostName of my ODI database. My target tables are located in a third server, in the same domain as my ODI database, but there are no DBLinks between them so I have to use SQL to SQL Append.
Thanks in advance.
Luiz Araujo (a.k.a. Buchu_VP)
Edited by: Luiz Araujo (a.k.a. Buchu) on Jun 1, 2010 7:38 AM -
Air for Android - Sqlite - Slow inserts
Hello there,
What is the best way to follow in order to make a lot of insert on a sqlite db?
I ask because I noticed a great and unpredicatable slowdown on my Galaxy S when I make a lot of insert.
I use an openAsync connection because I need to monitor what is happening.
Everytime a row is created I increment a counter.
It starts well, then after a 100 insert it slow down and nearly dies.
Any suggestions?
Thanks in advanceI optimized the query:
- I reuse the same statement and changing only the parameters
- I use the transaction with begin and commit
This speeded up the whole things in an incredible way, however I still notice a slow down that luckily, after the optimizations I made, never dies.
Any other suggestion will be appreciated! -
Slow inserts into partitioned table
I am having trouble inserting into a simple partitioned table after an upgrade to 11.2.0.3. I'm seeing insert speeds of subsecond up to 10 and 12 seconds. We have pre created the partitions for this table (and all children via reference partitioning). We have gathered dictionary and static object stats as well as statistics on all partitions.
Queries against the dictionary are incredibly slow as well and showing very high io.
Any help would be greatly appreciated. Thank you for your time.
Windows 2008 advanced server
Oracle Enterprise edition 11.2.0.3
Edited by: user593549 on Mar 26, 2012 11:16 AMuser593549 wrote:
I am having trouble inserting into a simple partitioned table after an upgrade to 11.2.0.3. I'm seeing insert speeds of subsecond up to 10 and 12 seconds. We have pre created the partitions for this table (and all children via reference partitioning). We have gathered dictionary and static object stats as well as statistics on all partitions.
Queries against the dictionary are incredibly slow as well and showing very high io.
Any help would be greatly appreciated. Thank you for your time.
Windows 2008 advanced server
Oracle Enterprise edition 11.2.0.3
Edited by: user593549 on Mar 26, 2012 11:16 AMThread: HOW TO: Post a SQL statement tuning request - template posting
HOW TO: Post a SQL statement tuning request - template posting -
Slow Insert APPEND into Temporary Table
Hello,
We did the following test on Oracle11g 11.1.0.7 database:
create global temporary table test_tab
as select * from tab1 where rownum <= 1;
insert into test_tab select * from tab1 where rownum <= 500000;
commit;
Elapsed time: 00:00:04.56
Statistic
80 recursive calls
26360 db block gets
10606 consistent gets
4729 physical reads
2543400 redo size
399 bytes sent via SQL*Net to client
340 bytes received via SQL*Net from client
6 SQL*Net roundtrips to/from client
1 sorts (memory)
0 sorts (disk)
500000 rows processed
truncate table test_tab;
insert /*+ append */ into test_tab select * from tab1 where rownum <= 500000;
commit;
Elapsed time: 00:00:09.35
Statistic
84 recursive calls
4900 db block gets
4738 consistent gets
4698 physical reads
1128 redo size
376 bytes sent via SQL*Net to client
354 bytes received via SQL*Net from client
6 SQL*Net roundtrips to/from client
1 sorts (memory)
0 sorts (disk)
500000 rows processed
Note that insert APPEND generates much less redo size: 1128 vs 2543400 . Now the question: why insert APPEND into the temporary table is two times slower than the ordinary insert? Potentially, it should run faster because of lower redo size... Any ideas?If you run the truncate / insert / truncate / insert append again are the timings the same ?
Wondering if the temp tablespace needed to extend for the first append. -
Detecting if insert or modify a footnote
Hi
I want to detect in my plugin if insert or modify a footnote on current document.
If user insert or modify a footnote, I will do something in my plugin.
How can I do?
thanksYou can use following code.
FUNCTION zcmbalance.
""Local Interface:
*" IMPORTING
*" REFERENCE(DATE) TYPE ZFINTBCM017-GLDATE
*" REFERENCE(COMPANY_CODE) TYPE ZFINTBCM017-GLCCODE
*" REFERENCE(GL_ACCOUNT) TYPE ZFINTBCM017-GLACCOUNT
*" REFERENCE(PROFIT_CENTER) TYPE ZFINTBCM017-GLPRFCENTER
*" REFERENCE(SUBSIDIARY_BALANCE) TYPE ZFINTBCM017-SUBBALANCE
TABLES:
ZBALANCEINT.
<b>DATA: ZBAL LIKE ZBALANCEINT.</b>
ZBAL-ZDATE = DATE
ZBAL-ZCODE = COMPANY_CODE
ZBAL-ZACCOUNT = GL_ACCOUNT
ZBAL-ZCENTER = PROFIT_CENTER.
SELECT SINGLE *
FROM ZBALANCEINT
WHERE ZDATE = DATE
AND ZCODE = COMPANY_CODE
AND ZACCOUNT = GL_ACCOUNT
AND ZCENTER = PROFIT_CENTER.
IF sy-subrc <> 0.
<b>ZBAL-ZBALANCE = SUBSIDIARY_BALANCE.</b>
<b>INSERT ZBALANCEINT FROM ZBAL.</b>
ELSE.
<b>ZBAL-ZBALANCE = SUBSIDIARY_BALANCE.</b>
<b>MODIFY ZBALANCEINT FROM ZBAL.</b>
ENDIF.
COMMIT WORK.
ENDFUNCTION. -
hi ,
i have a query that is fast in selection but somewhat very slow with the insertion with either create table as select OR the insert into table statment
my table is an empty table and does not have any indexes created
pls advise
tks & rgdshi ,
this is my SQLs
INSERT INTO /*+ APPEND */ AfterSales_Tbl
select a.txn_date , a.product_id , a.sales_date, a.sales_amountfrom MV_Txn a , MV_Sales bwhere a.txnid = b.salesidand b.sales_dates between trunc(sysdate) and trunc(sysdate) - 7
[/pre
rgds
Maybe you are looking for
-
How do i transfer photos from an old library to my current one ??????
how do i transfer photos from an Iphone that has been sync to an old Itunes library and transfer the data to my new itunes library ???????
-
FCPx is repeatedly crashing when I try to create a new project.
This started a few days ago, when FCPx crashed in the midst of my work, and I was unable to open the program without the spinning wheel. I deleted all events and projects, and re-installed the program, and now I can get it open successfully. But nwo
-
How do set multi ip address on solaris8
I set multi ip address by following command: inconfig hme0:1 159.226.140.55 netmask 255.255.255.0 system error:not found interface hme0:1. I shoud how to set multi ip address on solaris8
-
FONTS on OSX 10.4 And Under Classic (System 9 )
Okay - Ya' all be gentle. Remember - there are no STUPID questions. I'm running a new G5 but have lots of earlier stuff and a very large file of FONTS on my old - very old - G3 - system 9.2. Classic seems to be running fine but the fonts available in
-
Hi, I've read a lot of threads on ram upgrades, but still have a very basic question. I want to buy the $599 Mac Mini, which comes with 1G RAM. Clearly, the Mac RAM is overpriced, but I've been told by a friend to upgrade to at least 2G for $50 extra