Slow response from database

Database is suddenly slow. As a user can I see if any tables are locked? Any major operation going on?

> Database is suddenly slow.
Why the database? Maybe it is I/O that is slow? (someone copying/moving/deleting large files on the db server)
Maybe there is insufficient CPU capacity on the db server? (someone running 50 Perl processes concurrently to parse a bunch of files)
Maybe it is the network that is slow? (someone downloading movies and impacting the network segment that the db server is on)
> As a user can I see if any tables are locked?
Why locked tables? Oracle does not have lock escalation. Oracle row locks never escalates into page locks which in turn escalates to table locks.
> Any major operation going on?
Maybe.. but then it does not need to be a major operation. It could be that the CPU is running 1% busy, the I/O channels running at 1% capacity, plenty of free memory available, very little swapping.. and still, things seems "slow" despite nothing really being pushed even slightly hard on the server.
This scenario can be caused by many processes all waiting (in a queue) to access a single dedicated serialised/non-shared resource. And as the resource can only service one customer at a time, it is a major bottleneck - with the rest of the resources on that db server being grossly underutilised.
Mere observation ito measuring performance and determining performance problems is useless. The database may seem slow. But that does not in any way imply that the database itself is actually slow.
You have to look at things like the network, CPU, I/O and memory utilisation on the db server, to determine what component (if any) is responsible for decreasing performance.
At the same time, you also need to look at the wait states of the processes on that db server as maybe there is no resource problem, but a problem with serialisation and processing all waiting in a queue to be serviced.
Performance Tuning starts with The Identification of The Problem. So you need to pop the hood, on the network, disks, operating system and db server, to isolate and identify the problem.
Unfortunately, there is no easy magic wand to wave to uncover the performance issue.

Similar Messages

New OIF installation - Very slow response from OHS

Hi,
I've installed two instances of OIF 10.1.4.0.1 in order to create a test configuration to test XASP Attribute Sharing (i.e., we are not doing any other profiles).
Both OIF instances are using a separate instance of Sun Directory Server as its User Data Store, and both are using in-memory for Transient Data Store and have no Federation Data Store. Both OIF instances were installed similarly, using the "BASIC" installation choices, and are running on separate Redhat Linux servers.
The problem that we're seeing is very slow (10+ seconds to 30+ seconds) responses, even when we try to access a static HTML test page that we've placed in the OHS htdocs directory.
We've eliminated any networking problems as possibilities by a couple of tests that we've done, including shutting down the OHS and running a small, different test web server listening on port 7777. When we do that, response time for requests from that web server is almost instantaneous.
We are also seeing the same slow response when we try to access the test HTML page from the same server itself.
We have an SR open with Oracle, but so far, we've not been able to identify what the problem might be, so I'm hoping that perhaps someone else may have encountered this problem and have a resolution.
It is really puzzling because, as I said, this is a pretty vanilla default OIF installation, and it is doubly puzzling because even getting a tiny static HTML page is taking 10+ seconds. Also, I had installed OIF 10.1.4.0.1 on a Windows machine previously, and I don't see this problem on that earlier installation.
If anyone here has any insight into this problem, please let me know.
Thanks,
Jim

Hi,
I think I was able to temporarily eliminate this slow response problem. I think that the slow response problem is related to another problem that I posted about in the "Application Server - General" forum (see below, and Port 7200 conflict?
Jim
======= POSTED to Application Server - General forum ============
I think that I was able to temporarily work around this problem, plus another problem that I think is related (see: New OIF installation - Very slow response from OHS
What I did was to just comment out the "include" line for "dms.conf" in the Apache httpd.conf.
Doing that got rid of the "Address already in use" errors I was seeing in Apache error_log and also eliminated the slow response I was seeing from Apache/OHS (see that other post in "Identity Management" forum), BUT I am getting new errors about Apache not being able to find "htdocs/dms/Spy0".
I have the impression that dms is some kind of process monitoring system that Oracle has in 10gAS infrastructure? So, I have some questions:
1) What's the effect of dms not working?
2) Is it possible to just disable dms?
I hope that Oracle support can come up with a "real" fix for both problems, but in the meantime, as long as what I did doesn't break anything else besides dms, we can at least test the OIF stuff.
Jim

Slow response from SAR command

Hi,
I see slow response from SAR command on M5000 series server.
SunOS prod452 5.10 Generic_144488-09 sun4u 01/03/2012
sar 1 100
14:25:43 %usr %sys %wio %idle
14:25:45 1 1 0 98
14:25:46 1 0 0 99
14:25:48 1 1 0 98
14:25:50 1 2 0 97
14:25:52 1 1 0 98
14:25:54 1 1 0 98
14:25:56 1 2 0 97
14:25:58 1 2 0 98
14:26:00 1 2 0 97
14:26:01 1 2 0 97
Ideally i should have output at every one second, but it shows at every 2 seconds. Even i have tried to increase the interval but still i see it takes extra 1 sec. to give output.
These are the pretty newly configured Machine, do you think something to do with server perf./config/ etc ? Does this impacts perf. to my running workload on this machine ?
REgards
Tilok

Well, you could simply have a gigazillion fonts on your system that mess with the text engine. in that case some serious weeding out of fonts that you never use would help immensely. Otherwise you can always turn off hardware acceleration in the preferences, assuming the issue is related to that...
Mylenium

Slow response from search

When I type in the search bar, I get a very slow response--the letters appear sluggishly, one by one, with about a one second delay between each. I have about 6000 songs and they are all on an internal drive.
I created a new user account and dragged all my songs into iTunes in that account. Searching was lightning fast.
So it sounds like the problem is isolated to my account. I want to know if there is a way to speed up the search. The easy solution is to rebuild the xml file by deleting the songs from my library and re-importing them, but I do not want to lose my play counts, last played dates, etc. If there is a program that can save all that information for me, that would also be acceptable.
Any tips would be appreciated. Thanks!
--Tim
iMac G4 Mac OS X (10.4.6)

A few questions:
-Is the application running on the same server as the database?
-Single DB or RAC?
-When last was the database analyzed?(SQL Stats)
-Was there database admin scheduled jobs running(exports,RMAN ect.)
-When last did you rebuild indexes? (do we need to rebuild indexes)

To get list of suspended responses from database

Hi All,
we have O2A PIP installed with below versions:
Oracle AIA     11.1.1.5
Oracle WebLogic    10.3.5
SOA Suite    11.1.1.5
O2C PIP        11.1.1.1
In Production some of the orders are failing because of Siebel internal error and hence those responses are also erroring out in AIA.
Now as per the resequencing logic all subsequent responses for that account is going into suspended state so that only when the faulted response was recovered successfully, the subsequent response will flow.
The issue we have here is, whenever any new order is coming bearing same Account no. for which the response of some order has failed, the responses of new orders are also going into suspended state. now all suspended responses we can only see that in mediator_instance table, one row is getting added with same group_id of that account. leaving us with no option to identify if this response is for new order or same order which has faulted.
we can identify all the responses which has faulted by combining mediator_instance and composite_instance table as instance of updateSalesOrderSiebelComsProvABCSImpl gets created for the response which has failed.
but we need the list of all new orders and its group_id which is going into suspended state because of some other order's faulted response.
Please let us know if there is any way to get this list from database.
Thanks & Regards,
Vivek

In addition, this:
String filename = files;Should be this:
String filename = files;
That's just because he didn't use the "code" tags, so [ i ] made everything following it become italicized.

SQL Slow Response from remote DB

I have two sites of which I am executing SQL querys and updates to an Oracle DB. Both sites have the identical code. I am querying against an Oracle 9.0.2 DB. I have tried using both the 8i JDBC Driver and the 9i JDBC Driver. At one site which is closer to the DB site the response is 10 times as fast as the response from my second site. I've had the network administrators to test the network between sites for packet latency. The results were 40ms to pass packets between sites. Traceroute also showed the same response times. I have employed Oracle's assistance with tuning the database optimally. This has not improve the response time. I thought that the problem might be with my SQL so I created the following simple test and the response times were the same. Any thoughts as to were the problem might lie? Your help is much appreciated.
Site 1 response time: 4ms
Site 2 response time: 137ms
CODE:
try{
Class.forName("oracle.jdbc.driver.OracleDriver");
con=DriverManager.getConnection(url,name,pw);
if(con!=null){
System.out.println("Got Connection");
Statement s = con.createStatement();
long begintime = System.currentTimeMillis();
int bt = (int)begintime;
ResultSet rs = s.executeQuery("select sysdate from dual");
long endtime = System.currentTimeMillis();
int et = (int)endtime;
rs.next();
Time t = rs.getTime("sysdate");
System.out.println("Time: " + t.toString());
System.out.println("Time to Retrieve Records: "+ (endtime - begintime) + "ms");
rs.close();
con.close();
catch(SQLException sql){
System.err.println("SQL Error: " + sql.getMessage());
sql.printStackTrace();
catch(ClassNotFoundException cnfe){
System.err.println("SQL Error: " + cnfe.getMessage());

If it was me I would write a client server socket tester and run it on both machines. And do some tests that run individual and burst transmissions throughout the day. You could do the same thing without code by manually FTP'ing some large files between the systems. But that will have more of a real impact on the system however.
A simple trace route is only going to show a problem if it exists at a particular point in time.

Slow response from PE3

Sorry to be posting a question about this older version but I have just discovered the lack of Adobe support for PE3 and as a low volume user, I am not particularly keen to upgrade to PE7 if it won't recognize PE3 project files, and PE3 does all I want for now anyway. I am in the middle of editing this project by adding narration and background music, and finding very slow response to virtually everything I want to do. When trying to add narrative, it takes several minutes before it will start to record, then several more to save it. If I want to change any audio effects, same thing. Won't even open a function without a long delay, let alone save it. This problem only started as I've tried to add the audio. Up to now, editing the video clips has worked fine with only minimal delays.
The file is 9MB, which I wouldn't have thought to be a problem, and it's a reasonably new system, with 2GHz dual core AMD processor, 500GB of hard drive & 2GB of memory. Not sure what other info I should provide that might help. Has anyone else had this problem and or know what I should try to fix? very appreciative of any suggestions. TIA

Is yours a Vista machine? If so, you may want to optimize it for video editing.
http://www.adobeforums.com/webx/.3bc448a7
And, of course, ensure that you've got the 3.02 patch.
http://www.adobeforums.com/webx/.3bc2fd1f
The other thing that will greatly affect the program's performance is what you're using for source video.
The program works best with DV-AVIs files captured over FireWire from a miniDV camcorder. Other files (such as video from hard drive or DVD camcorders) pose a variety of challenges that can impact performance.
Another key issues is photographs. If you're using them as source files, you can easily choke the program if you're not optimizing them to no larger than 1000x750 pixels in size before you bring them into your project.
If you can offer more details, we can give you more specific advice.

Slow response from DB

Hi ,
My database is 10.02.0.4.0 on Solaris 5.9.
We are using Oracle 10g AS as Application server to
host the Application. Whenever the end user trying to access the
application who got the very solw response. At the application level
we have configured enough settings on datasource and connection pool parameters.
At Oracle Level 150 session, Processes 300 has set.
what could be the reason for End user getting problem like "Slow response","Connection Tiemd out".
Any parameter need to change. When I getting this error, what is the appraoch i need to follow to mitigate these kind of error in future .
Thanks,
Tippu

I have a performance checklist at :
Re: Multiple Sessions Performance Issue …………
Hth
Girish Sharma

Slow Response from the server

It seems that after 1 - 1:30 hrs after the server start the response from app server
(weblogic 5.1 here) is very very slow. This remains as it is until we stop and
restart the server..But again after the same period of time the problem continues
and the server tends to respond slowly.Where could be the problem? If you can
let us know the underlying problem, it would be great.

"Amit" <[email protected]> wrote in message news:<3e7054c6$[email protected]>...
It seems that after 1 - 1:30 hrs after the server start the response from app server
(weblogic 5.1 here) is very very slow. This remains as it is until we stop and
restart the server..But again after the same period of time the problem continues
and the server tends to respond slowly.Where could be the problem? If you can
let us know the underlying problem, it would be great.OK, check in the admin console for Garbage Collection. You will see
the heap usage reach 100% and the drop off suddenly. If this is the
case, consider increasing your heap, if that isnt an option.. Try
checking your code to make sure everything that is called is
destroyed.
That may help... If not it could be JDBC,other connectivity, or code.
But try that 1st and let me know.
Steve

Slow response from DPS

I am seeing a very slow response time from our Directory Proxy Server.
10:15:16 -0500 - PROFILE - INFO - conn=6 assigned to connection handler cn=default connection handler, cn=connection handlers, cn=config
10:15:36 -0500 - CONNECT - INFO - conn=6 client=127.0.0.1:48475 server=127.0.0.1:7389 protocol=LDAP
10:15:36 -0500 - OPERATION - INFO - conn=6 op=0 BIND dn="cn=proxy manager" method="SIMPLE" version=3 controls=""From the timestamps there is a 20 second delay (always seems to be the case) between when a connection request is handled by the "default connection handler" and the actual connection.
I am assuming the problem lies in a misconfiguration in the "default connection handler" .
I can provide additional info upon request.
Any suggestions would be greatly appreciated.

1 second for a query seems very slow...
Did you really measure 1s on the client side or did you get that number from the proxy logs ? Note that times shown in DPS logs may be inaccurate to some extends. For more info about that , have a look at
[http://blogs.sun.com/sduloutr/entry/etime_granularity|http://blogs.sun.com/sduloutr/entry/etime_granularity]

Getting response from Database by sending in parameters

Hi,
I am providing parameter which supplies string values to query the database using adapter. I can see the invoke activity sending the string value to the database, but respnse message does not contain anthing.
Can someone please tell me how to retrieve records from database by suplying parmeter value?
Thank You
Anil

hi Anil,
use a simple stored procedure or pl/sql function and specify the input and output parameters. We can send any no of parameters to a function and a function can able to return a value. so you can achieve your requirement using pl/sql functions. Learn about that more.
when you will define ParterLink (using DBAdapter browing databse to select your function ), you shud be able to view one OUT parameter with datatype what you have mentioned earlier in pl/sql function. When you define Invoke for that Parterlink, you should get Invoke Output variable with output variable (having name of that function.)
If you cannot browse the function in your DB Adapter, just make sure that the user you have defined in DB Adapter has enough privileges to access that function.
Regards
Rajesh A

Slow response from SRM 5.0

Hi Expert,
We have a recently upgraded SRM4.0 to SRM 5.0 server
The SRM application that we use via web-browser(ITS) is very slow.
Any approaches/suggestions to debug the slow response and improve on it.
Thanks
AK

Yes note 1318249 would be your best bet.
If you can access the same transactions in the SAPGui without performance problem then it would be a BASIS issue so you should alos look at BASIS notes.
Upgrage kernel first.
Hope this helps,
Regards,
Matthew

Null or empty response from database?

Is there anything I can do with connecting to my database? I get null or empty response. There is no error message.
package databaseconnection; import java.sql.Connection; import java.sql.DriverManager; import java.sql.ResultSet; import java.sql.SQLException; import java.sql.Statement; public class MainTest { static String userName = "mynickname"; static String userPassword = "myunforgettablepassword"; static String databaseUrl = "jdbc:mysql://databaseserver.com/myfavouritedatabase"; static String userQuery = "SELECT * FROM ques"; public static void main(String[] args) { // ConnectToDatabase CTD = new ConnectToDatabase(); MainTest MainTest= new MainTest(); MainTest.connect(userName, userPassword, databaseUrl, userQuery); // CTD.connect(userName, userPassword, databaseUrl, userQuery); System.out.println(); // i get null as the response? } void connect(String userName, String userPassword, String databaseUrl, String userQuery){ try { Class.forName ("com.mysql.jdbc.Driver").newInstance (); Connection conn = DriverManager.getConnection (databaseUrl, userName, userPassword); Statement stat = conn.createStatement(); String query = userQuery; ResultSet result = stat.executeQuery(query); try{ while (result.next()) { String s = result.getString("Name"); System.out.println(s); return; } } catch ( SQLException sqlException) { sqlException.printStackTrace(); } System.out.println("Result(s): "); while(result.next()){ System.out.println("name:\t" + result.getString("Name")); System.out.println(""); } conn.close(); }catch (SQLException e) { } catch (InstantiationException e) {} catch (IllegalAccessException e) {} catch (ClassNotFoundException e) {} } }
/* * Found here and modified from: * http://www.romow.com/computer-blog/how-to-use-mysql-with-java/ */ package databaseconnection; import java.sql.Connection; import java.sql.DriverManager; import java.sql.ResultSet; import java.sql.ResultSetMetaData; import java.sql.SQLException; import java.sql.Statement; import javax.swing.JOptionPane; public class ConnectToDatabase { String s; public void connect(String userName, String userPassword, String databaseUrl, String userQuery){ try { Class.forName ("com.mysql.jdbc.Driver").newInstance (); Connection conn = DriverManager.getConnection (databaseUrl, userName, userPassword); Statement stat = conn.createStatement(); String query = userQuery; ResultSet result = stat.executeQuery(query); try{ while (result.next()) { s = result.getString("Name"); System.out.println(s); return; } } catch ( SQLException sqlException) { sqlException.printStackTrace(); } System.out.println("Result(s): "); while(result.next()){ System.out.println("name:\t" + result.getString("Name")); System.out.println(""); } conn.close(); }catch (SQLException e) { } catch (InstantiationException e) {} catch (IllegalAccessException e) {} catch (ClassNotFoundException e) {} } }

Note: This thread was originally posted in the Swing forum, but moved to this forum for closer topic alignment.

Extremely slow response from Maverics with working account

Hi all,
let me reference an earlier issue I had after I switched to Mavericks on my MacBook:
https://discussions.apple.com/message/23576494?ac_cid=tw123456#23576494
I don't know if this problem is related to the earlier incident or not...
In any case, here is the scenario: MacBook (late 2008), previously with Mountain Lion. My main account ("nick") has a home folder location different than standard. This is because the boot drive is a (smallish) SSD, I.e. the home folder had been moved (long time ago) to a second HDD that I've installed instead of the Superdrive. The specialty here is that this 2nd HDD is encrypted (done at formatting time).
Subsequently, when I restart the MacBook, I first have to log into the admin account, which will automatically mount the HDD (password in keychain), then log back out and log into my user account.
This worked like a charm with Lion and Mountain Lion.
Along comes Mavericks. As you'll see in the link posted above, I had some issues getting the user to log in at all. Possibly a quirk in Mavericks, fixed by using a symbolic link.
In any case, the responsiveness of the system with that one account logged in is - to say the least - excessively slow.
It seems that something is holding up the CPU for long periods of time (30 seconds? 60 seconds?) between activities. This might be launching a program and waiting forever for it to show the initial window. Then selecting a function and waiting for that, etc.
Really odd is the fact that this only happens with this one account. I've tried logging in and then holding the Shift key (keeps the login script from loading selected programs automatically), but this didn't help.
Again, it only happens with this one account, so it has to be something to do with a setting or something.
Odd, too, is that the CPU isn't loaded at all - in fact, nothing seems to be happening (which, in fact, is what is happening...).
Ideas anyone?
I had a very similar effect when moving from Snow Leopard to Lion (Upgrade, not a clean install), so maybe it is something that isn't unusual?

Sorry to hear about the issues you've been having. Hard drives do sometimes fail, and it really isn't a fun experience when it does. I don't work for Apple, but I'm a certified tech that like others has access to the warranty repair system and I know how it works. Normally we get all parts very quickly, but there are some cases where backorders and incorrect repairs are inputted which slow things down.
Apple actually does read the forums and responds to them as well. I have several friends who do just that. I would suggest to the repair center that they try the "mark for review" checkbox because this shouldn't be taking this long and it will get a repair advisor's attention to find what's delaying it. Hopefully it'll be resolved shortly.

Slow response from oracle db

I am using application developed in Java. Database: oracle 10g & Webserver :Weblogic 9.2 for processing 10L records.
For this I have allocated 4GB ram to oracle & 2 GB to Weblogic. Also set 1000 sessions & 1000 process per users in oracle. Rollback segment is of 4GB.
Application uses multi threading for processing 10L records in Batch Job. During this processing Oracle server become slow. Even if maximum connection/sessions are less then 800, oracle does not allow application to create new connection because of this entire processing fails

As you are using the Oracle 10g and Webserver: Weblogic 9.2 on a single machine and the you have alloted only 4GB RAM to Oracle and 2 GB RAM to Weblogic. I strongly beleive that you should assign a minimum of 9GB RAM to Oracle and 5 GB RAM to Weblogic Server 9.2 edition in order to process 1000 Sessions and 1000 Process per users in Oracle. I one more thing I can say confidently that Oracle 10g and Weblogic should be managed on different dumps i mean to say on different machines. As they both use lot of Memory to run 1000 sessions and 1000 process per user you may require at 8GB RAM for running Oracle 10g and 3 GB to RAM to run Weblogic 9.2 server. And I believe that you may have to purchase lots of hardware stuff such as 3 processors (which run with 4 core technology). Please try with this configuration and let me know.

Slow response from database

Similar Messages

Maybe you are looking for