What are solutions for a way-too-big database?

Hi guys!
I'm a software developer and not very good in database designing. One day, I was asked something like this :
"For example, there is a company with a web application. One day, the database for that application is way too big, caused performance issues and others, what is the solution for that application database?"
At first, I thought that was about using multiple database with single app. But I don't know if I was right.
I want to ask that what are the solutions? If it's "multiple database" then what should I do? Using two connection to 2 database simutaneously?
I appreciate any replies. Thanks!

847617 wrote:
Thanks Lubiez Jean-Val... for your links.
I've got some more advices like :
- "transferring workload to another database using different techniques to copy the data from original db"
- "redesign of the database"
So that means we use 2 different databases?Sometimes it is deemed desirable to keep only fairly recent data on the OLTP database, where the normal transaction activity happens, and replicate the data to another database that also contains historical data. This second database is used for heavy reporting tasks.
And "redesign"?As in, design it from scratch and do it right this time. Make sure all data relations are properly defined to Third Normal Form; make sure all data is typed properly (use DATE columns for dates, NUMBER columns for numbers, etc); make sure you have designed effective indexing; make sure you use the capabilities of the rdbms and do NOT just use it as a data dump.
See http://www.amazon.com/Effective-Oracle-Design-Osborne-ORACLE/dp/0072230657/ref=sr_1_3?s=books&ie=UTF8&qid=1301257486&sr=1-3
are they really good solutions?Like most everything else, "It depends"
It depends on if the proposed solutions are implemented properly and address the root problem. The root problem (or even perceived problem) hasn't yet been defined. You've just assumed that at some undefined point the database becomes "way-too-big" and will cause some sort of problem.
It's assumed that we don't have or can't use partitioning.
And why is that assumed? Yes, you have to have a version of Oracle that supports it, and it is an extra cost license. But like everything else, you and your management have to do a hard-nosed cost/benefit analysis. You may think you can't afford the cost of implementing partitioning, but it may be that you can't afford the expenses derived from NOT implementing it. I don't know what the case is for you, but you and your management should consider the factors instead of just rejecting in out of hand.
:):)...You are making me - a student so excited about the history. From slides rule to the moon....
Edited by: 847617 on Mar 27, 2011 10:01 AMEdited by: EdStevensTN on Mar 27, 2011 3:24 PM

Similar Messages

All fonts in the toolbars are way too big after unplugging my laptop from a projector

When I plugged in a projector to my laptop (2010 MacBook Pro) all icons & fonts on my laptop went huge! I was able to mostly restore the size of the icons & fonts on my desktop after I unplugged the projector, but I can't get the font size in my toolbars back to the normal, standard size. I've tried to reset to default settings, but it makes no difference.

Although the guide is a little dated, a lot of the information there is really great and helps you track down a lot of files that are just taking up way too much space on your hard drive.
http://pondini.org/OSX/DiskSpace.html
I just did this the other day on my Air running 10.9 and recovered around 50GB of disk space, so there is potentially a lot you can do. I suggest following the steps on this guide (to the T, don't want to mess anything up) and go from there. Don't delete ANYTHING that you are not aware of the function, and remember that searching for names of files is a great way to determine whether or not you need it. Some files that SEEM like junk and take up a considerable amount of space are system files that will either be replaced the next time you boot up or render you computer useless entirely.
For example, there is a file that saves the state of your computer just before it goes to sleep so that it can return quickly when you start up again, and that file is normally the size of the amount of RAM you have (I have 4GB of RAM, so the file was 4GB). After doing some research, this file will either be recreated if deleted, or your sleep mode will not be able to function properly if the file is deleted. So again, just tread lightly.
Also, there is an application I used to get rid of a bunch of duplicate files that I had on my drive called Gemini (http://macpaw.com/gemini), which I recommend using to clean up documents. Again, be careful, you don't want to go deleting files that you are using that the application mistook for a duplicate.
Hope this helps!

Imported pictures are HUGE. Way too big. How do I reduce JPEG size?

JPEG size imports WAY TOO BIG. How do I reduce JPEG size? Thanks so much@!!

Set your camera to take lower quality pictures is the obvious answer here. You can reduce the size of image already in iPhoto when exporting them.

The "other" memory section on my Mac is WAY too big.

Hey everyone!
I know there are a LOT of discussions about this topic, but none of them really seemed to work for me. So the other day I got an error saying "Your startup disk is almost full". I immediately checked the storage section of "About my Mac", and sure enough, something like 100 GB were taken up by local backups. To fix this issue, I disabled local backups via this terminal command: sudo tmutil disablelocal.
I did a restart and checked my storage again. This time, it looked like this (and still does):
All of what used to be backups is now other. And the reason I have those 900 or so MB of backups is because I eventually turned those on again (with the reverse of the above command) in an effort to fix this. Out of options, I then went to the apple store, where they told me to just delete stuff that was too big. So I got a 1 TB drive and started copying and deleting stuff. The main thing I got rid of was a giant windows VM I had. And just to prove that something's wrong here, look at the breakdown of my files from OmniDiskSweeper:
If you do the math, I should have about 100 GB more free than I do now. I've tried:
Turning the thing off then on (you'd be surprised how often that works)
Resetting the PRAM
Forcing spotlight to reindex the machine
Throwing the machine out the window (well, not quite)
What I think might be the cause:
This is a stab in the dark here, but possibly all those GB of backups tried to delete themselves and somehow corrupted and aren not called backups anymore by the system?
Just one more thing: I noticed that OmniDiskSweeper showed a volume I don't recognize. It's called MobileBackups. Maybe that has something to do with it?
I'll appreciate your help.
Thanks!

Although the guide is a little dated, a lot of the information there is really great and helps you track down a lot of files that are just taking up way too much space on your hard drive.
http://pondini.org/OSX/DiskSpace.html
I just did this the other day on my Air running 10.9 and recovered around 50GB of disk space, so there is potentially a lot you can do. I suggest following the steps on this guide (to the T, don't want to mess anything up) and go from there. Don't delete ANYTHING that you are not aware of the function, and remember that searching for names of files is a great way to determine whether or not you need it. Some files that SEEM like junk and take up a considerable amount of space are system files that will either be replaced the next time you boot up or render you computer useless entirely.
For example, there is a file that saves the state of your computer just before it goes to sleep so that it can return quickly when you start up again, and that file is normally the size of the amount of RAM you have (I have 4GB of RAM, so the file was 4GB). After doing some research, this file will either be recreated if deleted, or your sleep mode will not be able to function properly if the file is deleted. So again, just tread lightly.
Also, there is an application I used to get rid of a bunch of duplicate files that I had on my drive called Gemini (http://macpaw.com/gemini), which I recommend using to clean up documents. Again, be careful, you don't want to go deleting files that you are using that the application mistook for a duplicate.
Hope this helps!

What are the down sides with very big SGA?

Hi
I have a heavy 9.2.0.8 64-bit database with very high dataloading and reporting having 24G SGA, 80 dual core CPU solaris 9. I would like to figure out what are the down sides of very big SGAs, and/or if increases the SGA.
Thanks

user5797895 wrote:
My question was not for a particular performance issue. I am thinking in different possibilities to improve to the best performance. I have statspack configured and we do tuning according to the stats.
Are you able to anlyze few spreport and give me your comments?Usually you have one or more particular issues that are worth to look into, e.g. particular standard reports that take longer than expected, batch runs that don't complete within the given timeframe, end-users complaining about ad-hocs reports taking too long, transactions that start to time out during peak workloads etc. etc.
If you're just interested in someone checking the current state/health of your system it's probably worth a try to post the most important parts of some statspack reports that you determine to be most representative for your system load. There are definitely some contributers here participating that should be able to provide some valuable insights.
I think it's always good to remember that the most significant performance improvement of a system is achieved by letting it do less work resp. avoid unnecessary work. So if you're e.g. able to reduce the logical I/O performed by the top n SQLs that put the most workload on your system significantly you usually gain much more than by just adding more and more memory/CPU to your system. Of course this is not the case if this kind of tuning has already been performed thoroughly but there is usually much room for improvement in this field.
Regards,
Randolf
Oracle related stuff blog:
http://oracle-randolf.blogspot.com/
SQLTools++ for Oracle (Open source Oracle GUI for Windows):
http://www.sqltools-plusplus.org:7676/
http://sourceforge.net/projects/sqlt-pp/

Service/Database Accounts - NT SERVICE\MSSQLSERVER & NT SERVICE\SQLSERVERAGENT - what are they for?

Hi Guys,
I’ve done a fair amount of research for this question but just cannot seem to find the answer to my question in simple, non-DBA, terms.
Server 2008 R2
SQL 2008 R2
There are 2 users in the system database logins (NT SERVICE\MSSQLSERVER, NT SERVICE\SQLSERVERAGENT) … what are they for? It appears that they
are accounts to run the corresponding Windows services but yet they cannot be selected from the list of available built-in accounts, local accounts or domain accounts.
Also, I am using a couple of domain user accounts to run the services, do I need to add them to the database? I changed the service accounts from NETWORK
SERVICE to the domain user accounts using the SQL Configuration Manger which is supposed to take care of managing the user group membership and registry changes but the domain accounts are not in the database …. The services appear to be running fine.
Thanks

In basic terms:
As you say, in the SQL Server Database Engine there are two logins;
NT SERVICE\MSSQLSERVER and NT SERVICE\SQLSERVERAGENT. The Database Engine runs in Windows as a Windows service named
MSSQLSERVER. The NT SERVICE\MSSQLSERVER login is used by the service to connect to the Database Engine. Basically, this is how it connects to itself. The SQL Server Agent runs as a Windows service named
NT SERVICE\SQLSERVERAGENT. The NT SERVICE\SQLSERVERAGENT
login is how the Windows process that is SQL Server Agent connects to the Database Engine to read the
msdb database to find out what it should do; and then do it. Both of these logins are members of the
sysadmin fixed server role, so they can do anything in the Database Engine. And they need to stay that way.
No, they can't be selected in the list of available built-in accounts, local accounts or domain accounts. This is because they are services, not accounts. They have a security identifier (SID) in Windows,
but Windows knows they aren't real users. Windows can authenticate them, but they don't have passwords that any human can use. If you run
lusrmgr.msc and look at the groups, you will see groups like
SQLServerMSSQLUser$computername$MSSQLSERVER and NT SERVICE\MSSQLSERVER
is a member of the group.
As for the account that you used to run the services, this is complicated and has changed from SQL Server 2005 to SQL Server 2008 and now again in SQL Server Code Named 'Denali'. The short answer is that
the account you specify will be used when a process tries to reach outside of the current Windows environment. But within the computer, there is a mix of authorization granted to the domain user, the service, and the Windows group
SQLServerMSSQLUser$computername$MSSQLSERVER.
The good news is that SQL Server Configuration Manager figures out all the stuff you need when you change the accounts. If you are a glutton for punishment, you can get an idea for how complicated this
is by looking at the Denali documentation where I have tried to provide more specific information. (Note this is not the same as SQL Server 2008.) You can see it at:
Configure Windows Service Accounts and Permissions
http://msdn.microsoft.com/en-us/library/ms143504(SQL.110).aspx
Rick Byham, Microsoft, SQL Server Books Online, Implies no warranty

What are solutions to the "Unexpected error while attempting to bind. Operation cancelled." message when trying to bind

What are solutions to the "Unexpected error while attempting to bind. Operation cancelled." message when trying to bind and i alreeady cheack my DNS settings and everything

Has this worked before, or is this a new configuration?
On the server, launch Terminal.app from Applications > Utilities and post the output of the following harmless diagnostic command — and you'll need to enter an administrative password when requested:
sudo changeip -checkhostname
On the OS X (presumably) Mavericks client, please post the IP address(es) of the DNS server(s) you're referencing. These are available via the System Preferences > Network > Advanced > DNS settings.
To check the console logs, launch Console.app from Applications > Utilities on both the client and the server, and then attempt the binding, and then post a (short! no more than ~25 to 50 lines!) of any errors related to the binding error.
If you're not in a position to post configuration details here — which is perfectly reasonable, and entirely your prerogative — or if I'm getting too confusing or too technical with some of the questions here, which would be my error of course — then (assuming these systems are under warranty) I'd suggest contacting the folks at the local Apple Store and having a chat with the local Geniuses about this, or checking directly with the Apple Support folks via telephone.

Issue syncing artwork: on-device DB is WAY too big!

I have an iPad and an iPhone. Both get music from the same source. Both take the tracks from my computer's library, downsize them to 128kps and there we are. However, the on-device Art Database sizes are very different and not in the way that makes sense. The size of the DB files was gathered using iPhone Explorer.
iPad has 4225 tracks. Art database is 2408MB
iPhone has 2334 tracks. Art database is 3104MB
A vast majority of the iPad library and the iPhone library overlap, so there is ZERO reason why the iPhone art DB takes up more space than the iPad's, unless something else is stored there that I dont understand. There are more FILES in the iPad DB, but they are overall smaller files. Each iPhone DB file is 200-500 MB. iPad DB files can be between 12-500MB
See the problem? Any thoughts? Any idea where I could direct this problem to, even? I just don't get it.
Message was edited by: J.Melrose

Although the guide is a little dated, a lot of the information there is really great and helps you track down a lot of files that are just taking up way too much space on your hard drive.
http://pondini.org/OSX/DiskSpace.html
I just did this the other day on my Air running 10.9 and recovered around 50GB of disk space, so there is potentially a lot you can do. I suggest following the steps on this guide (to the T, don't want to mess anything up) and go from there. Don't delete ANYTHING that you are not aware of the function, and remember that searching for names of files is a great way to determine whether or not you need it. Some files that SEEM like junk and take up a considerable amount of space are system files that will either be replaced the next time you boot up or render you computer useless entirely.
For example, there is a file that saves the state of your computer just before it goes to sleep so that it can return quickly when you start up again, and that file is normally the size of the amount of RAM you have (I have 4GB of RAM, so the file was 4GB). After doing some research, this file will either be recreated if deleted, or your sleep mode will not be able to function properly if the file is deleted. So again, just tread lightly.
Also, there is an application I used to get rid of a bunch of duplicate files that I had on my drive called Gemini (http://macpaw.com/gemini), which I recommend using to clean up documents. Again, be careful, you don't want to go deleting files that you are using that the application mistook for a duplicate.
Hope this helps!

What are the various methods/ways of assigining Cost Center to Org Units.

Hi,
What are the various methods/ways of assigining Cost Center to Org Units.
Regards
Rajesh

Hi ,
In PPOME tcode search for the org unit you have created.
On the Account Assignment tab page, relate your new organizational unit to a cost center in the detail area. Select the Master Cost Center field, and search for a cost center using the search help F4
Regards,
Malathi V

Safari font specifically in Google search is way too BIG

The font in Google search is way too big. Only in Google. Other websites show normal fonts. Did someone encounter such problem? Please help me! Thank you!

Hello How Goes It,
select the cells you want to have no margin and choose the text inspector of the infopalette and the text tab button. At the bottom is a slider and a number field. There you can adjust the margin thickness.
Frank.

My company loaded profiles onto my iPad for email and calendars.. There is also a signing certificate and a certificate. What are these for? Additionally are they able to monitor apps and usage, ie Internet usage when it is not on their wifi?

My company loaded profiles onto my iPad for email and calendars.. There is also a signing certificate and a certificate. What are these for?
Additionally are they able to monitor apps and usage, ie Internet usage when it is not on their wifi?
I do not have any VPN enabled?

Do you happen to have an Android? If so and depending on what version there is a great data usage analyse tool built-in. See if you can go to Settings -> Data Usage from there you can pick a current or previous billing cycle and then use the vertical sliders to select a date range and it will filter the usage data per app to show you exactly what app(s) were using data during that time frame.

What are attributes for a row?

What are attributes for a row?
Are these the column names associated with the row?

Attribute is the object-oriented term for an object's data holders. So attributes are analogous to columns in a table.
Of course, attribute also has a couple of non-OO meanings. So the answer to your question really depends upon context.
Cheers, APC

What are alternatives for passing DB credentials to rwrun/rwclient command line instead of USERID parameter?

What are alternatives for passing DB credentials to rwrun/rwclient command line instead of USERID parameter?
As per application security, the DB details should be mentioned in command line.
I am looking for alternatives which I can use?
Please help.
Regards
Gaurav Anand

If you want a database username/password for every user that runs a report, you need to include un/pw in the command.
An option that you see quite often is that Reports use a general database user to run any report. In that case you can include the un/pw of this special user in the Key Map File:
http://docs.oracle.com/cd/E17904_01/bi.1111/b32121/pbr_run013.htm#RSPUB23557
Of course, in that case you need an application in front of the call to the report to check if the user is allowed to run the report. This is how most web applications work anyway: there is one database user, and application users are maintained in the application itself.
Or you can use single sign on:
http://docs.oracle.com/cd/E24269_01/doc.11120/e24479/pbr_sso002.htm#RSPUB23512

What the steps are useful for us while designing the database by Performanc

Dear Experts
Plz tell me.
What the steps are useful for us while designing the database by Performance point of view.
Like Proper indexing and table assing to specific tbalespace etc

Oracle manuals would do the best for you. There is no step by step rules for tuning the database , had this be true , then Oracle must've write a programme , application or sql script to achive it . I would advise you to go through oracle performance and tuning guide .
Hare Krishna
Alok

What are rule for adding names to photos?

what are rule for adding names to photos in iphoto?

Since your profile indicates you have a Mac, I am guessing you are asking how to add a face name using iPhoto on your Mac.
You would open the photo in iPhoto, click "Info" on the bottom, and then the "Faces" tab on the right.
Once you have added the face name in iPhoto on your Mac, once you sync the photo to your iPhone, the name will appear there as well.

What are solutions for a way-too-big database?

Similar Messages

Maybe you are looking for