How to investigate DB performance degradation.

We use Oracle11gr2 on win2008R2.
I heard that DB performance degradation is happening and I would like to know how to improve DB performance.
How could I investigate the reason of DB performance degradation ?

Hi,
the first thing to establish is the scope of the problem -- whether it's the entire database, a single query, or a group of queries which have something in common. You cannot rely on users for that.
Then depending on the scope of the problem, you pick a performance tool that matches the scope of the problem, and use it to obtain the diagnostic information.
If you can confirm that the issue is global (almost everything is slow, not just one query), then AWR and ASH may be helpful. For local (i.e. one or several queries) issues, you can use SQL trace, dbms_xplan and ASH. Keep in mind that ASH and AWR require a Diagnostic and Tuning Pack license.
Best regards,
Nikolay

Similar Messages

Performance degradation with addition of unicasting option

We have been using the multi-casting protocol for setting up the data grid between the application nodes with the vm arguments as
*-Dtangosol.coherence.clusteraddress=${Broadcast Address} -Dtangosol.coherence.clusterport=${Broadcast port}*
As the certain node in the application was expected in a different sub net and multi-casting was not feasible, we opted for well known addressing with following additional VM arguments setup in the server nodes(all in the same subnet)
*-Dtangosol.coherence.machine=${server_name} -Dtangosol.coherence.wka=${server_ip} -Dtangosol.coherence.localport=${server_port}*
and the following in the remote client node that point to one of the server node like this
*-Dtangosol.coherence.wka=${server_ip} -Dtangosol.coherence.wka.port=${server_port}*
But this deteriorated the performance drastically both in pushing data into the cache and getting events via map listener.
From the coherence logging statements it doesn't seems that multi-casting is getting used atleast with in the server nodes(which are in the same subnet).
Is it feasible to have both uni-casting and multi-casting to coexist? How to verify if it is setup already?
Is performance degradation in well-known addressing is a limitation and expected?

Hi Mahesh,
From your description it sounds as if you've configured each node with a wka list just including it self. This would result in N rather then 1 clusters. Your client would then be serviced by the resources of just a single cache server rather then an entire cluster. If this is the case you will see that all nodes are identified as member 1. To setup wka I would suggest using the override file rather then system properties, and place perhaps 10% of your nodes on that list. Then use this exact same file for all nodes. If I've misinyerpreted your configuration please provide additional details.
Thanks,
Mark
Oracle Coherence

Performance degradation factor 1000 on failover???

          Hi,
          we are gaining first experience with WLS 5.1 EBF 8 clustering on
          NT4 SP 6 workstation.
          We have two servers in the cluster, both on same machine but with
          different IP adresses (as it has to be)!
          In general it seems to work: we have a test client connecting to
          one of the servers and
          uses a stateless test EJB which does nothing but writing into weblogic.log.
          When this server fails, the other server resumes to work the client
          requests, BUT VERY VERY VERY SLOW!!!
          - I should repeat VERY a thousand times, because a normal client
          request takes about 10-30 ms
          and after failure/failover it takes 10-15 SECONDS!!!
          As naive as I am I want to know: IS THIS NORMAL?
          After the server is back, the performance is also back to normal,
          but we were expecting a much smaller
          performance degradation.
          So I think we are doing something totally wrong!
          Do we need some Network solution to make failover performance better?
          Or is there a chance to look closer at deployment descriptors or
          weblogic.system.executeThreadCount
          or weblogic.system.percentSocketReaders settings?
          Thanks in advance for any help!
          Fleming


See http://www.weblogic.com/docs51/cluster/setup.html#680201
          Basically, the rule of thumb is to set the number of execute threads ON
          THE CLIENT to 2 times the number of servers in the cluster and the
          percent socket readers to 50%. In your case with 8 WLS instances in the
          cluster, add the following to the java command line used to start your
          client:
          -Dweblogic.system.executeThreadCount=16
          -Dweblogic.system.percentSocketReaders=50
          Hope this helps,
          Robert
          Fleming Frese wrote:
          > Hi Mike,
          >
          > thanks for your reply.
          >
          > We do not have HTTP clients or Servlets, just EJBs and clients
          > in the same LAN,
          > and the failover should be handled by the replica-aware stubs.
          > So we thought we need no Proxy solution for failover. Maybe we
          > need a DNS to serve failover if this
          > increases our performance?
          >
          > The timeout clue sounds reasonable, but I would expect that the
          > stub times out once and than switches
          > to the other server for subsequent requests. There should be a
          > refresh (after 3 Minutes?) when the stub
          > gets new information about the servers in the cluster, so he could
          > check then if the server is back.
          > This works perfectly with load balancing: If a new server joins
          > the cluster, I automatically receives
          > requests after a while.
          >
          > Fleming
          >
          > "Mike Reiche" <[email protected]> wrote:
          > >
          > >It sounds like every request is first timing out it's
          > >connection
          > >attempt (10 seconds, perhaps?) on the 'down' instance
          > >before
          > >trying the second instance. How do requests 'failover'?
          > >Do you
          > >have Netscape, Apache, or IIS with a wlproxy module? Or
          > >do
          > >you simply have a DNS that takes care of that?
          > >
          > >Mike
          > >
          > >
          > >
          > >"Fleming Frese" <[email protected]> wrote:
          > >>
          > >>Hi,
          > >>
          > >>we are gaining first experience with WLS 5.1 EBF 8 clustering
          > >>on
          > >>NT4 SP 6 workstation.
          > >>We have two servers in the cluster, both on same machine
          > >>but with
          > >>different IP adresses (as it has to be)!
          > >>
          > >>In general it seems to work: we have a test client connecting
          > >>to
          > >>one of the servers and
          > >>uses a stateless test EJB which does nothing but writing
          > >>into weblogic.log.
          > >>
          > >>When this server fails, the other server resumes to work
          > >>the client
          > >>requests, BUT VERY VERY VERY SLOW!!!
          > >> - I should repeat VERY a thousand times, because a normal
          > >>client
          > >>request takes about 10-30 ms
          > >>and after failure/failover it takes 10-15 SECONDS!!!
          > >>
          > >>As naive as I am I want to know: IS THIS NORMAL?
          > >>
          > >>After the server is back, the performance is also back
          > >>to normal,
          > >>but we were expecting a much smaller
          > >>performance degradation.
          > >>
          > >>So I think we are doing something totally wrong!
          > >>Do we need some Network solution to make failover performance
          > >>better?
          > >>Or is there a chance to look closer at deployment descriptors
          > >>or
          > >>weblogic.system.executeThreadCount
          > >>or weblogic.system.percentSocketReaders settings?
          > >>
          > >>Thanks in advance for any help!
          > >>
          > >>Fleming
          > >>
          > >

Performance Degradation - High fetches and Prses

Hello,
My analysis on a particular job trace file drew my attention towards:
1) High rate of Parses instead of Bind variables usage.
2) High fetches and poor number/ low number of rows being processed
Please let me kno as to how the performance degradation can be minimised, Perhaps the high number of SQL* Net Client wait events may be due to multiple fetches and transactions with the client.
EXPLAIN PLAN FOR SELECT /*+ FIRST_ROWS (1) */ * FROM SAPNXP.INOB
WHERE MANDT = :A0
AND KLART = :A1
AND OBTAB = :A2
AND OBJEK LIKE :A3 AND ROWNUM <= :A4;
call     count       cpu    elapsed       disk      query    current        rows
Parse      119      0.00       0.00          0          0          0           0
Execute    239      0.16       0.13          0          0          0           0
Fetch      239   2069.31    2127.88          0   13738804          0           0
total      597   2069.47    2128.01          0   13738804          0           0
PLAN_TABLE_OUTPUT
Plan hash value: 1235313998
| Id | Operation                    | Name   | Rows | Bytes | Cost (%CPU)| Time     |
|   0 | SELECT STATEMENT             |        |     2 |   268 |     1   (0)| 00:00:01 |
|* 1 | COUNT STOPKEY               |        |       |       |            |          |
|* 2 |   TABLE ACCESS BY INDEX ROWID| INOB   |     2 |   268 |     1   (0)| 00:00:01 |
|* 3 |    INDEX SKIP SCAN           | INOB~2 | 7514 |       |     1   (0)| 00:00:01 |
Predicate Information (identified by operation id):
   1 - filter(ROWNUM<=TO_NUMBER(:A4))
   2 - filter("OBJEK" LIKE :A3 AND "KLART"=:A1)
   3 - access("MANDT"=:A0 AND "OBTAB"=:A2)
       filter("OBTAB"=:A2)
18 rows selected.
SQL> SELECT INDEX_NAME,TABLE_NAME,COLUMN_NAME FROM DBA_IND_COLUMNS WHERE INDEX_OWNER='SAPNXP' AND INDEX_NAME='INOB~2';
INDEX_NAME      TABLE_NAME                     COLUMN_NAME
INOB~2          INOB                           MANDT
INOB~2          INOB                           CLINT
INOB~2          INOB                           OBTAB
Is it possible to Maximise the rows/fetch
call     count       cpu    elapsed       disk      query    current        rows
Parse      163      0.03       0.00          0          0          0           0
Execute    163      0.01       0.03          0          0          0           0
Fetch   174899     55.26      59.14          0    1387649          0     4718932
total   175225     55.30      59.19          0    1387649          0     4718932
Misses in library cache during parse: 0
Optimizer mode: ALL_ROWS
Parsing user id: 27
Rows     Row Source Operation
28952 TABLE ACCESS BY INDEX ROWID EDIDC (cr=8505 pr=0 pw=0 time=202797 us)
28952   INDEX RANGE SCAN EDIDC~1 (cr=1457 pr=0 pw=0 time=29112 us)(object id 202995)
Elapsed times include waiting on following events:
Event waited on                             Times   Max. Wait Total Waited
----------------------------------------   Waited ---------- ------------
SQL*Net message to client                  174899        0.00          0.16
SQL*Net more data to client                155767        0.01          5.69
SQL*Net message from client                174899        0.11        208.21
latch: cache buffers chains                     2        0.00          0.00
latch free                                      4        0.00          0.00
********************************************************************************

user4566776 wrote:
My analysis on a particular job trace file drew my attention towards:
1) High rate of Parses instead of Bind variables usage.
But if you look at the text you are using bind variables.
The first query is executed 239 times - which matches the 239 fetches. You cut off some of the useful information from the tkprof output, but the figures show that you're executing more than once per parse call. The time is CPU time spent using a bad execution plan to find no data -- this looks like a bad choice of index, possibly a side effect of the first_rows(1) hint.
2) High fetches and poor number/ low number of rows being processedThe second query is doing a lot of fetches because in 163 executions it is fetching 4.7 million rows at roughly 25 rows per fetch. You might improve performance a little by increasing the array fetch size - but probably not by more than a factor of 2.
You'll notice that even though you record 163 parse calls for the second statement the number of " Misses in library cache during parse" is zero - so the parse calls are pretty irrelevant, the cursor is being re-used.
Regards
Jonathan Lewis
http://jonathanlewis.wordpress.com
http://www.jlcomp.demon.co.uk
To post code, statspack/AWR report, execution plans or trace files, start and end the section with the tag {noformat}{noformat} (lowercase, curly brackets, no spaces) so that the text appears in fixed format.
"Science is more than a body of knowledge; it is a way of thinking"
Carl Sagan

Performance degradation. Need advice on starting all over again.

I have never formatted my drive or reinstalled my OS X like I used to do with XP during the windows days. Now there is some performance degradation and opening of applications like Safari, iPhoto and others is really slow. I maintain Time Machine backups of my full Snow Leopard partition. What should I do? Format the HD, re install SL, or simply restore from TM or reinstall and then restore...?
I dont really want to carry my windows attitude to mac, of reformatting, reinstalling and then starting all apps from scratch. I wanna leverage out of my TM backup. Please advice.
Neerav
MacBook 2.4GHz Unibody (Late 2008), 2GB RAM, SL

The hatter wrote:
Those steps, repair permissions? only checks the installed application receipts -- worthless.
Disk Utility doesn't check for bad blocks, and Apple First Aid misses and doesn't fix directory problems that are picked up by 3rd party tools like Disk Warrior.
The hatter's comments do not represent a consensus of opinion about this & are at least partially misleading.
Permissions repairs are indeed limited to comparing receipt info to actual permissions settings, but that is hardly worthless. It is well documented that mis-set permissions will cause a number of problems & resetting them to receipts values is an effective cure for those specific problems. Obviously, that won't cure problems with other causes, but since there is no magic cure-all it would be foolish to expect it to behave like one.
Regarding Disk Utility, it is true that it can't repair certain problems that some 3rd party utilities can; however, it is very effective at identifying file system problems, including those for some file systems the 3rd party apps do not support. It is also the most conservative disk utility available, designed not to attempt any repair that could result in loss of data. This is one reason it isn't as powerful as the 3rd party ones -- it is best to use it first if you suspect you have file system problems & use the more powerful ones only when necessary.
To be fair, Disk Warrior includes a directory optimization function that Disk Utility doesn't. However, an "unoptimized" directory isn't a problem in & of itself, & it is debatable how much real world benefit there is to optimizing the directory, at least with the current OS & modern high performance drives. I used to see noticeable improvements by periodically using Disk Warrior with OS 9 & the drives of that era, but these days my Macs & Snow Leopard seem to do just fine without it.
Basically, it is simple: use the tool that best does what you need to do. There is no benefit from using a sledge hammer when a tack hammer will do; in fact, the sledge hammer may do more harm than good, or just wear you out for no good reason. Also consider the wisdom of the old saying that to a hammer everything looks like a nail. Sometimes, you don't need a tool at all, just the wisdom to know that you don't.
Regarding bad sectors, every drive has them. That is not a concern by itself but the drive suddenly developing new ones is a sure sign of serious problems. Drives keep track of this themselves. Utilities provide a way to query the drives about this & may provide early warning of impending failure, but since the drive is providing the info this is not 100% reliable. For this reason, whether you use one or not, it is extremely important to backup your important data to other devices regularly & often.

Performance degradation using Jolt ASP Connectivity for TUXEDO

We have a customer that uses Jolt ASP Connectivity for TUXEDO and is suffering
from a severe performance degradation over time.
Initial response times are fine (1 s.), but they tend to increase to 3 minutes
after some time (well, eh, a day or so).
Data:
- TUXEDO 7.1
- Jolt 1.2.1
- Relatively recent rolling patch installed (so no there are probably no JSH performance
issues and memory leaks as fixed in earlier patches)
The ULOG shows that during the night the JSH instances notice a timeout on behalf
of the client connection and do a forced shutdown of the client:
040911.csu013.cs.kadaster.nl!JSH.234333.1.-2: JOLT_CAT:1185: "INFO: Userid:
[ZZ_Webpol], Clientid: [AP_WEBSRV3] timed out due to inactivity"
040911.csu013.cs.kadaster.nl!JSH.234333.1.-2: JOLT_CAT:1198: "WARN: Forced
shutdown of client; user name 'ZZ_Webpol'; client name 'AP_WEBSRV3'"
This happens every 10 minutes as per configuration of the JSL (-T flag).
The customer "solved" the problem for the time being by increasing the connection
pool size on the IIS web server.
However, they didn't find a "smoking gun" - no definite cause for the problem.
So, it is debatable whether their "solution" suffices.
It is my suspicion the problem might be located in the Jolt ASP classes running
on the IIS.
Maybe the connection pool somehow loses connections over time, causing subsequent
users having to queue before they get served (although an exception should be
raised if no connections are available).
However, there's no documentation on the functioning of the connection pool for
Jolt ASP.
My questions:
1) What's the algorithm used for managing connections with Jolt ASP for TUXEDO?
2) If connections are terminated by a JSH, will a new connection be established
from the web server automatically? (this is especially interesting, because the
connection policy can be configured in the JSL CLOPT, but there's no info on how
this should be handled/configured by Jolt ASP connectivity for TUXEDO)
Regards,
Winfried Scheulderman

Hi,
For ASP connectivity I would suggest looking at the .Net client facility provided in Tuxedo 9.1 and later.
Regards,
Todd Little
Oracle Tuxedo Chief Architect

Performance degradation in pl/sql parsing

We are trying to use xml pl/sql parser and noticed performance degradation as we run multiple times. We zeroed into the following clause:
doc := xmlparser.getDocument(p);
The first time the procedure is run the elapsed time at sqlplus is something like .45sec, but as we run repeatedly in the same session the elapsed time keeps on increasing by .02 seconds. If we log out and start fresh, we start again from .45sec.
We noticed similar degradation with
p := xmlparser.newParser;
but we got around by making the 'p' variable as package variable, initializing it once and using the same for all invocations.
Any suggestions?
Thank you.

Can I enhance the PL/SQL code for better performance ? Probably you can enhance it.
or, is this OK to take so long to process these many rows? It should take a few minutes, not several hours.
But please provide some more details like your database version etc.
I suggest to TRACE the session that executes the PL/SQL code, with WAIT events, so you'll see where and on what time is spent, you'll identify your 'problem statements very quickly' (after you or your DBA have TKPROF'ed the trace file).
SQL> alter session set events '10046 trace name context forever, level 12';
SQL> execute your PL/SQL code here
SQL> exitWill give you a .trc file in your udump directory on the server.
http://www.oracle-base.com/articles/10g/SQLTrace10046TrcsessAndTkprof10g.php
Also this informative thread can give you more ideas:
HOW TO: Post a SQL statement tuning request - template posting
as well as doing a search on 10046 at AskTom, http://asktom.oracle.com will give you more examples.
and reading Oracle's Performance Tuning Guide: http://www.oracle.com/pls/db102/to_toc?pathname=server.102%2Fb14211%2Ftoc.htm&remark=portal+%28Getting+Started%29

Performance degradation after setting filesystemio_option=setall from none.

Hi All,
We have facing performance degradation after setting filesystemio_option=setall from none on my two servers as mentioned below.
Red Hat Enterprise Linux AS release 4 (Nahant Update 7) 2.6.9 55.ELhugemem (32-bit)
Red Hat Enterprise Linux Server release 5.2 (Tikanga) 2.6.18 92.1.10.el5 (64-bit)
We are seeing lots of Disk I/O happening. We expected "*filesystemio_option=setall* " will improve performance but it is degrading. We getting slowness complains.
Please let me know do we need to set somethign else along with this ...like any otimizer parameter( e.g. optimizer_index_cost_adj, optimizer_index_caching).
Please help.

Hi Suraj,
<speculation>
You switched filesystemio_options to setall from none, so, the most likely reason for performance degradation after switching to setall is the implementation of directio. Direct I/O will skip the filesystem buffer cache, and and allow Oracle to read directly from disk to the database buffer cache. However, on a system where direct I/O is not implemented, which is what you had until you recently messed with that parameter, it's likely that you had an undersized database buffer cache, but that was ok, because many (most) of the physical I/Os your database was doing, were actually being serviced by the O/S filesystem buffer cache. But, you introduced direct I/O, and wiped out the ability of the O/S to service any physical I/Os from filesystem buffer cache. This means that every cache miss on the database buffer cache, turns into a real, physical, spin-the-disk, move-the-drive-head, physical I/O. And, you are suffering the performance consequences.
</speculation>
Ok, end of speculation. Now, assuming that what I've outlined above is actually going on, what to do? Why is direct I/O lower performing than buffered, non-direct I/O? Shouldn't it's performance be superior?
Well, when you have an established system that's using buffered I/O, and you switch to direct I/O, you almost always will have to increase the size of the database buffer cache. The problem is that you took a huge chunk of memory away from the the O/S, that it was using to buffer your I/Os and avoid physical I/O. So, now, you need to make up for it, by increasing the size of the database buffer cache. You can do this, without buying more memory for the box, because the O/S is no longer going to need to use so much memory for filesystem buffers.
So, what to do? Is it worth switching? Well, on balance, it makes sense to use direct I/O, and give Oracle a larger database buffer cache, for the simple fact that (particularly on a server that's dedicated to being an Oracle database server), Oracle has far more sophisticated caching algorithms, and a better understanding of the various types of data being cached, and so should be able to make more efficient use of the memory, than the (relatively) brain dead caching algorithms of the kernel and filesystem mechanisms.
But, once again, it all comes down to this:
What problem are you trying to solve? Did you have any I/O related issues? Do you have any compelling reason to implement direct I/O? Rule #1 is "if it ain't broke, don't fix it." Did you just violate rule #1? :-)
Finally, since you're on Linux, you can use the 'free' command to see how much memory is on the box, how much is free, and how much is dedicated to filesystem cache buffers. This response is already pretty long, so, I'm not going to get into details, however, if you're not familiar with the command, the results could be misleading. Read the man page, and try to be clear about understanding it before you make any assumptions about the output.
Hope that helps,
-Mark

Can System performance degraded by Z development

Hi,
Can system performance degrade by more z development? Is there any idle
ratio with standard system/reports to Z development in % terms?
Regards,
Shail

Z programs are just programs not written by SAP, but added by the customer.
The performance of the Z programs depends on the experience of the Z programmer.
There is no ideal ratio, I just depends how much must be programmed different from the standard.
And oyu should also keep in mind, that if a Z-program replaces a standard program that it will have broader/narrower functionality, so it will need a different amount of time.
There is no optimum defined, you must find out whether something can be done better.
Siegfried

How to mesure/benchmark performance of a new database on new server?

Hi there
I have two oracle servers with following (same) details:
RHEL 5.8 64-bit
Oracle 10gR2 - 10.2.0.5.8
ASM 10gR2 - 10.2.0.5.8
Server A: RAM 32GB, 8 CPUs @ 3.00GHz
Server B: RAM 128GB, vCPUs 16 cores
Server A (physical server) already has a database A. Server B (on VMWare - yes, my client is moving all Oracle servers to VMware for whatever reason) is a new installation and new database B with exact same init params as databas A. I expdp the data from database A and impdp into database B.
As per the hardware team, the hardware is better than the old server B. I did a very basic test to check if new DB performs better than that on physical server. Here is the results:
I ran a simple query to create a new table. The original table (say, table_a) contains 1.7+ million rows and size is 2.2GB.
create table test1
as
select * from table_a;
It took 3:28mins on database B while it took only 1:55mins on database A. So the new database B seems to be performing poor (apparently). Then I looked at the explain plan (not sure if it means much because it s a very simple query) and here it is from both databases:
Database A (physical server
Plan
SELECT STATEMENT ALL_ROWS
Cost: 14,052 Bytes: 2,161,302,003 Cardinality: 16,250,391
1 TABLE ACCESS FULL TABLE table_a
Cost: 14,052 Bytes: 2,161,302,003 Cardinality: 16,250,391
Database B (virtual server)
Plan
SELECT STATEMENT ALL_ROWS
Cost: 59,844 Bytes: 2,161,302,003 Cardinality: 16,250,391
1 TABLE ACCESS FULL TABLE table_a
Cost: 59,844 Bytes: 2,161,302,003 Cardinality: 16,250,391
Questions:
1. Why is the cost different? Should I "compute statistics" on database B (virtual server)?
2. How to investigate further and find out reason for the time difference?
3. What other benchmark test can I run to make sure that I have the right database configuration?
Not sure if this is enough info - if not, please let me know what else should I provide.
The team I have to hand-over this server is refusing to accept it by saying that it is slower than the existing one.
Please advise!
Best regards

Wow... I am really thankful for everyone's input - this is really really appreciated!
I will try what you all have suggested. In the meantime, I did some simple test on both databases and here are the results:
Create table t1
(1.7million rows)
Create index on
two columns on t1
Create table t2
500000 rows
Create Index on
two columns on t2
Delete from t1
(500000 rows)
Insert into t1
500000 rows
Drop table t2
Drop table t1
Database A
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
1st run
00:01:55.78
00:02:12.59
00:00:03.06
00:00:01.99
00:01:25.56
00:00:10.37
00:00:00.15
00:00:05.12
2nd run
00:01:56.27
00:02:11.54
00:00:02.89
00:00:01.09
00:01:18.39
00:00:10.20
00:00:00.17
00:00:04.87
3rd run
00:01:56.71
00:02:12.36
00:00:03.14
00:00:01.13
00:01:22.97
00:00:10.22
00:00:00.15
00:00:04.88
Database B (VM)
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
hh:mm:ss.00
1st run
00:00:25:83
00:03:54.60
00:00:00.67
00:00:01.43
00:00:29.56
00:00:09.75
00:00:00.05
00:00:01.10
2nd run
00:00:24.67
00:03:05.81
00:00:00:62
00:00:01.10
00:00:31.76
00:00:08.59
00:00:00.04
00:00:00.59
3rd run
00:00:44.06
00:03:12.91
00:00:00.97
00:00:01.62
00:00:39.35
00:00:08.90
00:00:00.03
00:00:00.61
Now, the database on Server B (VMware) seems to be outperforming that on Server A except for the "Create index on two columns on t1" column.
Any clues why index creation is consistently taking longer on the database B (on VM) as compared to that on database A (physical server)?
@jgarry: I am not in a position to try SLOB (no doubt a good tool with lots of reputation) because it requires to create a new DB (which I cannot do on the existing server). I did try "HammerDB" but unfortunately it crashed on each attempt to test the load.

SQL Performance Degrades Severely in WAN

The Oracle server is located in the central LAN and the client is located in the remote LAN. These two LANs are connected via 10Mbps wide network. Both LANs are of 100M bps inside. If the SQL commands are issued in the same LAN as Oracle server is located in, the speed is fast. However, if the same commands are issued in the remote LAN, the speed is degraded severly, almost 10 times slower. There are only a few results returned by these SQL commands. My questions are the reason of this performance degradeness and how to improve the performance in the remote client.
The server is Oracle817 and OPS, and the SQL commands are issued in PB programs in the remote client.
Thanks very much.

Thank you very much.
I found another point which might lead to the performance problem. The server's Listener.ora is configured as following:
LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC))
And the client's TNSNAMES.ORA is configured as following:
EMIS02.HZJYJ.COM.CN =
(DESCRIPTION =
(ADDRESS_LIST =
(ADDRESS = (PROTOCOL = TCP)(HOST = 172.26.17.18)(PORT = 1521))
(CONNECT_DATA =
(SERVICE_NAME = emis)
It shows that the listener protocol is set as IPC. However, the client is set to the protocol of TCP. Would there be a network latency for the protocol conversion between IPC and TCP?
Thanks a lot.

Database performance degrade - delete operation

Hi,
I have a big database. One of the table contains 120 milion records and many tables (more than 50) has referential integrity to this table.
Table structure
Customer (Cust_ID, and other columns). Cust_ID is Primary Key. Other tables has referential integrity to Customer.Cust_ID.
There are around 100 thousand records that have entry only in this (Customer) table. These records are identified and kept in a table Temp_cust(Cust_ID).
I am running a PL/SQL which fetches a Cust_ID from Temp_cust and delete that record from Customer.
It is observed that a delete command takes long time and the whole system performance degrades. Even a on-line service that inserts row into this table looks like almost in hand stage.
The system is 24/7 and I have no option to disable any constraint.
Can someone explain why such a simple operation degrades system performance. Please also suggest how to complete the operation without affecting performance of other operations.
Regards
Karim

Hi antti.koskinen
There is no on delete rule. All are simple
referential integrity.
Like REFERS CUSTOMER (Cust_ID).
Regards,
KarimCan you run the following snippet just to make sure (params are name and owner of the Customer table):
select table_name, constraint_name, constraint_type, delete_rule
from dba_constraints
where r_constraint_name in
       select constraint_name
       from dba_constraints
       where owner = upper('&owner')
       and table_name = upper('&table_name')
       and constraint_type = 'P'
/Also check the last time the table was rebuilt - deletes w/o rebuilds tend to raise the high water mark.

SCOM reports "A significant portion of the database buffer cache has been written out to the system paging file. This may result in severe performance degradation"

This was discussed here, with no resolution
http://social.technet.microsoft.com/Forums/en-US/exchange2010/thread/bb073c59-b88f-471b-a209-d7b5d9e5aa28?prof=required
I have the same issue. This is a single-purpose physical mailbox server with 320 users and 72GB of RAM. That should be plenty. I've checked and there are no manual settings for the database cache. There are no other problems with
the server, nothing reported in the logs, except for the aforementioned error (see below).
The server is sluggish. A reboot will clear up the problem temporarily. The only processes using any significant amount of memory are store.exe (using 53GB), regsvc (using 5) and W3 and Monitoringhost.exe using 1 GB each. Does anyone have
any ideas on this?
Warning ESE Event ID 906.
Information Store (1497076) A significant portion of the database buffer cache has been written out to the system paging file. This may result in severe performance degradation. See help link for complete details of possible causes. Resident cache
has fallen by 213107 buffers (or 11%) in the last 207168 seconds. Current Total Percent Resident: 79% (1574197 of 1969409 buffers)

Brian,
We had this event log entry as well which SCOM picked up on, and 10 seconds before it the Forefront Protection 2010 for Exchange updated all of its engines.
We are running Exchange 2010 SP2 RU3 with no file system antivirus (the boxes are restricted and have UAC turned on as mitigations). We are running the servers primarily as Hub Transport servers with 16GB of RAM, but they do have the mailbox role installed
for the sole purpose of serving as our public folder servers.
So we theroized the STORE process was just grabbing a ton of RAM, and occasionally it was told to dump the memory so the other processes could grab some - thus generating the alert. Up until last night we thought nothing of it, but ~25 seconds after the
cache flush to paging file, we got the following alert:
Log Name:      Application
Source:        MSExchangeTransport
Date:          8/2/2012 2:08:14 AM
Event ID:      17012
Task Category: Storage
Level:         Error
Keywords:      Classic
User:          N/A
Computer:      HTS1.company.com
Description:
Transport Mail Database: The database could not allocate memory. Please close some applications to make sure you have enough memory for Exchange Server. The exception is Microsoft.Exchange.Isam.IsamOutOfMemoryException: Out of Memory (-1011)
   at Microsoft.Exchange.Isam.JetInterop.CallW(Int32 errFn)
   at Microsoft.Exchange.Isam.JetInterop.MJetOpenDatabase(MJET_SESID sesid, String file, String connect, MJET_GRBIT grbit, MJET_WRN& wrn)
   at Microsoft.Exchange.Isam.JetInterop.MJetOpenDatabase(MJET_SESID sesid, String file, MJET_GRBIT grbit)
   at Microsoft.Exchange.Isam.JetInterop.MJetOpenDatabase(MJET_SESID sesid, String file)
   at Microsoft.Exchange.Isam.Interop.MJetOpenDatabase(MJET_SESID sesid, String file)
   at Microsoft.Exchange.Transport.Storage.DataConnection..ctor(MJET_INSTANCE instance, DataSource source).
Followed by:
Log Name:      Application
Source:        MSExchangeTransport
Date:          8/2/2012 2:08:15 AM
Event ID:      17106
Task Category: Storage
Level:         Information
Keywords:      Classic
User:          N/A
Computer:      HTS1.company.com
Description:
Transport Mail Database: MSExchangeTransport has detected a critical storage error, updated the registry key (SOFTWARE\Microsoft\ExchangeServer\v14\Transport\QueueDatabase) and as a result, will attempt self-healing after process restart.
Log Name:      Application
Source:        MSExchangeTransport
Date:          8/2/2012 2:13:50 AM
Event ID:      17102
Task Category: Storage
Level:         Warning
Keywords:      Classic
User:          N/A
Computer:      HTS1.company.com
Description:
Transport Mail Database: MSExchangeTransport has detected a critical storage error and has taken an automated recovery action. This recovery action will not be repeated until the target folders are renamed or deleted. Directory path:E:\EXCHSRVR\TransportRoles\Data\Queue
is moved to directory path:E:\EXCHSRVR\TransportRoles\Data\Queue\Queue.old.
So it seems as if the Forefront Protection 2010 for Exchange inadvertently trigger the cache flush which didn't appear to happen quick or thuroughly enough for the transport service to do what it needed to do, so it freaked out and performed the subsequent
actions.
Do you have any ideas on how to prevent this 906 warning, which cascaded into a transport service outage?
Thanks!

JNLP ClassLoader performance degraded in 1.6 significantly

JNLP ClassLoader performance degraded in 1.6 significantly. We have measured the performance in comparison to former JRE versions, which shows a significant performance problem
JAR Type               Java PlugIn Appletload-time
OriginalSize(13.7MB)     1.5.0_05     17:00
OriginalSize(13.7MB)     1.6.0_03     25:00
OriginalSize(13.7MB)      1.6.0_07     25:00
OriginalSize(13.7MB)     1.6.0_10     24:00
Please give a suggestion that how to improve performance of the application.

Cross-posting can frustrate anyone who tries to help you only to find out later that the same answer was given hours ago in a cross-posted thread. No one likes wasting their time, especially a volunteer. The polite thing to do would be to provide links in both cross-posts to each other.

How bad is the performance hit with RTMPT?

In a conversation with engineers recently at a CDN it was suggested to me that streaming all video over RTMPT only was a viable solution to the barriers posed by firewalls and proxies blocking port 1935. They indicated that they had seen no signifigant performance degradation with tunneling and that many of their clients were making the switch from "rollover" type connection models to connection via RTMPT only.
This ran counter to my notion of the process. I have always thought the packet overhead was signifigant.
Is it? How bad is the performance hit for streaming live h.264 video?

Hmmm... That's quite a hit.
I'm trying to determine the best strategy for reaching the most people with the least performance hit. A guy lays out his strategy here:
http://www.kensodev.com/2010/02/19/rtmp-being-blocked-by-firewalls-flash-media-server/
He basically says ditch 1935. Never use it. Always use 80. Like this:
rtmp://your_ip_address:80/app_name
If that fails, do this:
rtmpt://your_ip_address:80/app_name
Does that seem valid? Does that first option avoid the perfomance hit of tunnelling while getting you more connections? If so, it makes me think there is no benefit at all to connecting via 1935.

How to investigate DB performance degradation.

Similar Messages

Maybe you are looking for