Want to know The Truth About CPM?

12 April 2011

Yeah, yeah 11.1.2.1 is here, so what?

Introduction

Given the excitement around this release (geeks get enthused by some pretty odd things), you’d think an Oracle-red lightning bolt would shoot down from the heavens when I type that and smite me mightily like Thor’s hammer right on the noggin.  BONG!  Ouch, that hurt.  But I have an excuse for writing that.

I have the problem all (most) EPM implementers face – ask us to build an Essbase database, write a calc script, craft an MDX formula, proof of concept a Planning app – that we can do with aplomb and flair.  We do that for a living; we better be good at it.

But install a package that requires more than double-clicking on setup.exe and we (or at least I) go all wobbly.  I have spent countless and fruitless hours trying to get development environments to work since System 9.  Oh, eventually I mostly get them running but we are sometimes talking months (say hello System 9.0.1 and Oracle Fusion EPM 11.1.13 and 11.1.2) to get to a working state.  It is my no-longer-secret shame.  I suck at installations.  There, I have said it and I feel much better.

The problem is, I want to experience the joy that is 11.1.2.1.  Alas, my successful installation will likely occur sometime around the general release of 11.1.2.2.  What to do?

There is an answer to my (and your) problem

The answer’s name is John Booth.  He has very generously provided an Amazon Web Services (AWS) Amazon Machine Image (AMI) with the basics of 11.1.2.1 installed on it FOR FREE.

That means, with an AWS account, you (and I) can run 11.1.2.1 without installing anything.  Oh joy, oh rapture.  Maybe I can even have a life.  Maybe.  Maybe you can, too.

You can stop reading right now and go to his blog here to try out his EPM Test Drive to get all of the details.  Or keep on reading to see how someone even as technically lame as yr. obdnt. srvnt. can run 11.1.2.1 with practically no effort at all.

Sign up for Amazon Web Services

You must have an Amazon Web Services account.  Go here to start the process.  Once you have that, follow these steps to 11.1.2.1 nirvana.

Amazon has put together a nice tutorial.  I recommend that you go through it, following the Windows path as the 11.1.2.1 EPM Test Drive AMI is on Windows.  Even if you’re a Linux bigot, Go Team Windows!

Don’t be overwhelmed by all of the steps.  There is a bit of a learning curve but it isn’t that bad.

Launching 11.1.2.1 in the cloud

1)      Assuming you’ve signed up and run through the tutorial, find your IP address at http://www.whatsmyip.org/  You are going to need this to define your security access to the 11.1.2.1 image.
2)      Log onto Amazon Web Services at http://aws.amazon.com


3)      Username = youremailaddress.com, password = whateveritis.

4)      Make sure you have Adobe Flash 10.2 installed.  The below message only pops up if you don’t have Flash 10.2 installed.

5)      Click on the EC2, the Security Group, and finally the Create Security Group link.  You are going to need your IP address from step #1

6)      Name the group and give a description.  I am going to pretend my name is Alice and that she/I live in Philly, PA.  We will not talk about my identity crisis.
7)      In the bottom half of the screen, select RDP as a connection protocol:

You need only enter your will need your IP address with a /32 switch.  Then click on Add Rule.

8)      Change the IP address to:  youripaddress:32, e.g., 98.111.148.105/32 and then click on Save.  You will see the following as confirmation:

9)      Then click on Apply Rule Changes.

10)      Now click on the AMIs link.

11)   Select All Images, type in ami-b0669bd9 ami-de2bd4b7, and then click on Refresh.  The AMI should appear.

12)    Right click on the AMI and select Launch Instance

13)   Select m2.xlarge aka Extra Large and then Continue.  Don’t worry about the Availability Zone.  The power is almost yours.

14)   Take the default on the Advanced Instance Options.

15)   Take the default on Tags, or put in a value if you like.  It doesn’t really matter unless you managing multiple instances.

16)   If you haven’t already created a key pair, do so now.  Make sure that you keep that in a secure place as it is part of your authentication into AWS.  In this case I created one called “AmazonCloud”.

17)   Leave the security group at default and click on Continue.

18)   You will get a chance to review your selection.  Once you’re satisfied, click on Launch.

19)   You will receive a confirmation message.  Click on the View your instances on the Instances page link.

20)   You will see a list of running instances.  Now you play the waiting game as you are now subject to a delay equivalent to powering up a real server in your data center. 


21)   Once you’ve waited 10 to 15 minutes, right click on the AMI line, and select Connect.

22)   You will get information on how to connect to the instance.  Don’t be discouraged if you can’t; remember, it’s a virtual server oh-so-slowly booting off the drive.  You can either download the shortcut to your hard drive, or copy the public DNS to the clipboard.  Either way, you’re going to be running the Terminal Services client, aka, Remote Desktop, to connect to the server.

23)   If you click on the Download shortcut file link, save it to your Windows desktop for sanity’s sake.

NB – This is in FireFox; Internet Explorer’s dialog box looks a little different.


24)   Double click on the icon to see if it’s really running or not.

25)   You will get this scary-looking error message.  Ignore it, it’s much worse than it looks and click on Connect.

26)   Happy, happy, joy, joy.  You have now connected.
 
27)   Once started, check out how much power you have. 


28)   Go ahead, start EPM 11.1.21 by double-clicking on the Workspace shortcut .  Admit it, you’ve been waiting all day for this.  Now play, play, play or more seriously (and boringly – see, I just invented a word, my 11.1.2.1 exuberance knows no linguistic bounds) learn, learn, learn.

29)   When your brain cannot take any more, use the Start menu to Shut Down the server.  Remember, if the instance is running, you are getting charged by the hour.

30)   Your 11.1.2.1 instance is now stopped.  Any work that you have done is stored on the stopped hard drive.  If you are well and truly done with the instance of the virtual machine, right click on that Stopped AMI and select Terminate.  This will delete the AMI and its drives.  You lose all work.  If you want to keep the drive in the state you left it, leave it a Stopped state or make your own AMI from it and then Terminate the instance.  So long as the instance exists, even in a Stopped state, you are being charged, albeit a small amount.


31)    That’s it.  Have you sent John an note of appreciation yet?  :)

The best way to thank John is to come see our presentation at Kaleidoscope

You’ll have to tolerate me, as well.  Seriously, I come at this from an implementation perspective, John from an infrastructure perspective.  It’s the best of both worlds.

Think of us as Martin and Lewis.  Just call me Jerry; Dino’s the infrastructure guy.  Have a gander of us entertaining your spiritual ancestors at their convention.

96 comments:

Unknown said...

So basically you are saying your presentation at Kscope will be as funny as a Jerry Lewis comedy skit?

Cameron Lackpour said...

It won't be the Colgate Comedy hour, or The Errand Boy.

I'm saying I'll be the cringeworthy awkward one and John will be the smooth one.

OTOH, I'll get a telethon, so there's always that.

Regards,

Cameron Lackpour

P.S. Seriously, going to the cloud has been a big exercise in being out of my comfort zone. Sometimes, when I'm frustrated and confused, I want to run around yelling, "Hey you Mrs. Lady person" or just drooling out of the side of my mouth. I whine to John, he patiently listens, and then just does whatever I was having a problem with. Hence the whole Lewis = me, Martin = him analogy.

Nathan Low said...

Hi Cameron,

I've got everything set up on the AWS side, but when I log into the server, it doesn't look like EPM is set up yet. Were there any steps, like editing a host file, that needed to be done to get the EPM system up and running?

Thanks,
Nathan

Cameron Lackpour said...

Nathan,

Are you sure you launched the right AMI?

Can you even see Oracle EPM in the Start menu?

How about the services -- anything there?

Ping me on LinkedIn and we can set up a G2M session tonight to review your issue if you haven't resolved it by then.

Regards,

Cameron Lackpour

Nathan Low said...

Hi Cameron,

Yeah, I've got the correct AMI, I think - ami-b0669bd9.

EPM system is there, I can see it. I'm currently running the diagnostics to see what is happening. Ran the weblogic administration server for epmsystem1 to see what is happening, and I've started up all the application servers. I'm wondering if I need to set an IP address up somewhere?

Thanks,
Nathan

Nathan Low said...

Looks like the diagnostics say that nothing is configured. I guess I have to configure everything, then try it all out?

Cameron Lackpour said...

Nathan,

You don't have to do that. I don't know what's going on in the instance you launched, but you do *not* have to configure anything to get the basics of 11.1.2.1.

See my instructions -- all I did was launch the Workspace shortcut.

If you're having this much trouble, terminate the instance and relaunch the AMI in a new instance.

Again, if you can wait till after work (I am in the US East coast) I would be more than happy to do a GoToMeeting session with you to work out what's going on. Send me a LinkedIn message -- it comes through to my email.

I am sorry you are having trouble -- it is a bit much to try both 11.1.2.1 and AWS all at the same time but once you get your feet wet with AWS you'll learn to like it.

Regards,

Cameron Lackpour

Nathan Low said...

Whoops, fixed it. I needed to start SQL Server, then make sure everything is stopped, then run Oracle EPM System -> Foundations Services -> "Start EPM System".

This is pretty cool! Will check it out later.

Thanks for the help.

John A. Booth said...

Hi Nathan, you don't have to configure IP or SQL Server. I think you may have an underpowered machine you attached to the AMI. Try this with a m2.xlarge and wait about 5 minutes after starting it to login. You should then be able to immediately hit workspace, use EAS, and Essbase.

Regards,

John

Jagan Mohan Tulasi said...

Hi Cameron

I am not able to connect after doing all the steps it says contact your network administrator

Please advise

Thanks
JMT

Anonymous said...

John, any chance to have Office installed into the image to experiment Smart View as well?
Do you have plans to upgrade the image with Planning?
TIA

John A. Booth said...

Hi JMT,

Regarding Office as it's a licensed component it isn't on the image. You are free to upload and install your own office as once you have the AMI on your instances it's your machine.

On the planning front I have the databases created and it's a matter of about 10-15 minutes to enable planning. Would a walk through be useful for that?

Are you still having difficulties firing up the image? Where do you get that contact your network administrator message? Can you post something on google images?

Regards,

John

bwise said...

John and Cameron: Thanks! Great image.

I was able to get SmartView working from my local machine. Thought I might share the steps.

1. Open port 19000 (for shared connections) and 13080 (for private connections) on the Amazon firewall.

2. Try to access workspace at http://amazonIPaddress:19000/workspace/ This didn't work for me. You'll notice that Apache returned http://www.metavero.com:19000/workspace. This is caused by the ServerName directive in httpd.conf. I haven't tried changing this yet. Instead, I updated my local hosts file to resolve metavero to my Amazon IP address--but now I cannot get to metavero.com :(

3. Login to workspace at http://www.metavero.com:19000/workspace

4. Download & Install SmartView

5. Login to EAS and load some data into the sample cubes.

6. Open Excel and change my shared connection URL to http://www.metavero.com:19000/workspace/SmartViewProviders

7. Added a new server called metavero and start querying. Surprisingly snappy.

Anonymous said...

I signed up for AWS to try this out but I can't find the AMI. I searched among all AMIs but I don't see it. Am I missing something?

Cameron Lackpour said...

Anon,

I logged into http://console.aws.amazon.com, went to the EC2 home, and then went to the AMI link. I then typed in ami-de2bd4b7 in the search box and got...nothing. Whoops, I used "Owned by me" instead of "All Images" in the Viewing dropdown. Once I changed to All Images, the 11.1.2.1 AMI showed correctly. Take a look at the step 11 in my post and you'll see exactly how to set up the console.

Regards,

Cameron Lackpour

Cameron Lackpour said...

Bwise,

I thought for sure that your issue was that you didn't have port 80 opened up.

However, I spoke with John and he said:
>>He probably didn’t cycle the machine after opening the port.

>>80 shouldn’t really need to be opened up. 19000 will proxy everything behind the scenes.

Would you believe he's right again? I created a security group with 3389, 19000, and 13080 and then used that group to connect to his 11.1.2.1 image and I was able to connect to Essbase via SmartView on port 19000 AND to connect to Workspace via my laptop's browser.

What I *didn't* do was your trick with updating the local hosts file. I simply put in the Amazon DNS name, e.g., http://ec2-184-73-120-11.compute-1.amazonaws.com:19000/workspace/SmartViewProviders

The problem with the above is that every time I stop and then restart the instance (or relaunch from the AMI) the IP address will be changed.

I think John is going to highlight how to manage this in our KScope11 presentation. Perhaps you should attend our presentation?

:)

Regards,

Cameron Lackpour

Paul said...

Great stuff Cameron!
Question - have you tried to install ODI with this AWS instance? Compatibility matrix says it won't work with SQL Server 2008. So, could we install ODI locally and then hook up to the instance?

thanks,
Paul English

Cameron Lackpour said...

Paul,

Thanks for your kind words, but your thanks really ought to go to John Booth -- he set up the AMI -- all I did was write a tutorial.

In answer to your question re could you set up locally and connect to the server -- absolutely. You will need to open up the appropriate ports in the Security Group. I have no idea what ports ODI uses but that shouldn't be impossible to figure out. I will note that I used ODI remotely and found it to be very chattty and thus somewhat slow, but I can be pretty impatient.

If you are coming to Kaleidoscope, I have a section in my whitepaper that shows how to open ports to allow SmartView through to your client. I'll likely blog about that in future as well. If you figure out what ports ODI uses, I'll put that into the blog post as well.

Last thing -- if you want a good discussion of what OS works and what one does not with the different releases of ODI, check out this Network54 thread:
http://www.network54.com/Forum/58296/thread/1307020529/11.1.2.1+and+ODI

Regards,

Cameron Lackpour

John A. Booth said...

Hi Paul,

I have used ODI both 10.1.3.6 with SQL 2008. At this stage I would suggest using 11.1.1.5.0. The Microsoft JDBC drivers require Java 1.6 so in the case of 10.1.3 you need to:

1) Download Java 4 drivers from Microsoft.
2) Extract those and locate the jdbc driver.
3) Update ODI_JAVA_HOME to be a 1.6 java
4) Place the SQL jdbc 4.0 in path
5) In ODI use driver: com.microsoft.sqlserver.jdbc.SQLServerDriver
6) URL jdbc:sqlserver://metavero:1433;SelectMethod=cursor;databaseName=ODIMASTER;integratedSecurity=false
NOTE: Above presumes db name created was ODIMASTER.

Anonymous said...

I'm having trouble using RDP to log into the ami. This is the first time I've used it, so it's quite possible I missed something in the setup, but I also wanted to check if this is still up and running as expected. I am trying to connect to the following ami:
ami-de2bd4b7.

Any help is appreciated! Thank you.

Cameron Lackpour said...

Anon,

You do have the right AMI and yes, it's still there.

Were you able to tell if the AMI launched? You can do this by right clicking on the instance and selecting (I think) Get System Log. That will pop up a window showing whether Windows has booted. Sometimes that can take an awfully long time (like 10 minutes or so).

If you send me a message via LinkedIn, I'd be happy to set up a G2M session to walk through the process with you.

Regards,

Cameron Lackpour

yannis vossos said...

Cameron i cant find the instance , coming in from Brazil. Do you think that has anything to do?

Cheers

yannis

Cameron Lackpour said...

Yannis,

The only thing I can think of is that you aren't picking US East (Virginia) when you are searching for the AMI?

I don't believe that AWS filters access based on your geographic area.

Regards,

Cameron Lackpour

donworry said...

Hi,

I was searching for the AMI ami-b0669bd9 and was unable to find it. I was in the US East (Virginia) region during the search. Is it still available?

Also, I did try using AMi ami-de2bd4b7) which has a description of Oracle EPM 11.1.2.1 Image. When I started this instance I was never [6 hours] able to retrieve a adminstrator password. I noticed in your blog you didn't seem to have to generate or enter a password wehn you RDP'd to the server. Am I missing something?

Any help would be appreciated.

Thanks

Cameron Lackpour said...

Don (don't Worry, it's not worth the bother),

You are right, the AMI is ami-de2bd4b7 and I have amended the instructions accordingly. I think John must have changed the AMI after I wrote the tutorial. Or I was dreaming. Either one.

Re the password -- it's epmtestdrive. You only need to generate a password on AWS when you launch one of Amazon's generic Windows instances. As this AMI is already created, there's no need to do this. See John's blog post for all of the usernames and passwords.

Regards,

Cameron Lackpour

donworry said...

Thanks Cameron

Anonymous said...

following up on the last post from someone. I am putting into Remote Desktop

"Administrator"
"epmtestdrive"

also tried admin, adminstrator...

any ideas why it is not opening the instance?

Cameron Lackpour said...

Dear Anon,

Are you sure you launched the right AMI?

I just checked -- I was able to successfully log in to the AMI with administrator/epmtestdrive.

Regards,

Cameron Lackpour

Anonymous said...

I think so... I followed your instructions and when it came time to put in admin and password, did so with no luck to ami-de2bd4b7

The only thing I can think of is you mention a 15 min wait... When I start the instance, it says "pending" then "running" so I assume it is ready to connect. Is that correct?

Anonymous said...

If you right click on the instance and select "Get System Log" when the instance has started (and the times do seem to vary -- this is what happens when you don't own the infrastructure) you should see something like "Windows is read to use" and a bunch of other text showing an RDP certificate.

Frustratingly, AWS doesn't always return the above so I generally wait about five minutes and try, then 10 minutes, and finally 15 minutes. It almost always starts up at that point.

If you like, ping me on LinkedIn and I'll set up a G2Meeting with you and we can try it out together.

Regards,

Cameron Lackpour

Unknown said...

This is a great post. Thanks so much for making it available.

I get to step 26, but there is no Happy, happy, joy, joy. I get "This computer can't connect tot the remote computer".

How do I go about connecting? Contact Amazon? Are there logs to look at? Do I need to open some ports?

Cameron Lackpour said...

Marv,

First off, you have to thank John Booth -- I only wrote up the instructions he did all the work.

I use the AMI constantly for development work -- it is really rock solid.

As for your connection issue, it sounds like you didn't open port 3389/RDP. Also, if you did, and are trying this from work, it is more than just a little possible that your employer blocks port 3389. I know that when I go to client sites I can often get in through their guest wireless network but not on my client PC which is on the "real" network. Have you tried this from home?

Regards,

Cameron Lackpour

Unknown said...

Cameron --
Thanks for your post. You are exactly right. I am trying this from work, and apparently RDP is blocked. When I tried it from home it worked perfectly. Thanks for helping me understand this...
Marv

Anonymous said...

Doesn't anyone have a image like the old BICG 2009 Image Oracle had of 11.1.1.1?????? That way we can just open up in VMWare Workstation?

kw said...

Cameron or John,

My AMI 'State' says 'Running' and 'Status Checks' says '2/2 checks passed'. However when I click on the RDP link I downloaded to Desktop it doesn't connect?? I'm on my home machine not work. Do I have to configure anything else?

Cameron Lackpour said...

Anon,

If you are a partner, check out this link:
http://www.linkedin.com/groups/Oracle-EPM-Amazon-Web-Services-144765.S.88594763?view=&srchtype=discussedNews&gid=144765&item=88594763&type=member&trk=eml-anet_dig-b_pd-ttl-cn&ut=1syWHKnl9vCB41

You can get the latest and greatest from OPN.

And of course if you want to run that AMI on VMWare:
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1018015

Regards,

Cameron Lackpour

Cameron Lackpour said...

Sam,

Are you sure you have the ports for your security group opened up? This is my problem 99% of the time when I launch this from a new location.

Regards,

Cameron Lackpour

Bruno Andrade said...

Hi there,

Do you have an image already with Planning as well?

Cameron Lackpour said...

Bruno,

I believe the Planning install binaries are on the image but are not installed.

Regards,

Cameron Lackpour

kw said...

Yep. But how do you get to the configure stage? It's not the same as 11.1.1.3??? Are the repositories all SQL Server or Oracle? Also, I'm trying to convert to VMWare. How can I add another drive big enough to hold the vmdk files on the image???

Cameron Lackpour said...

Sam,

Go take a look at my KScope whitepaper from 2011 on the cloud. I show how to expand the size of an existing drive.

I am going to put it up as a blog post but I've been on a Special Project (which will be announced in the next few days) or I would put the drive expansion up as a post.

That whitepaper has some pretty (ahem) decent things in it, including the most complete guide to using the AWS API from the command line. It'll all go up on this blog, eventually, but that Special Project has to get its plug first.

I will say this -- I did my Special Project (when oh when will I end with the shilling? Soon, soon.) I did it 100% on the AWS cloud -- it worked perfectly.

Regards,

Cameron Lackpour

kw said...

I'm scrolling through your blog and don't see what your referring to. Do you have a link?

Cameron Lackpour said...

Sam,

I haven't put it up yet.

Go to www.odtug.com, then click on Tech Resources, then search for "cloud". The presentation and the whitepaper will pop up.

Or contact me via LI and I'll email it to you.

Regards,

Cameron Lackpour

Bruno Andrade said...

Yes, I can see the binaries there and if I try to install, it says that its already installed and when I try to run the configuration, it shows that nothing has been configured.

What should I do in this case? configure only planning or reconfigure everything?

thanks again

Cameron Lackpour said...

Bruno,

Take a look at www.metavero.com. John Booth put up a new AMI with Planning installed yesterday.

Regards,

Cameron Lackpour

kw said...

OUCH!!! I have been working for two days on getting the other one converted to VMWare... :)

kw said...

Bruno,

Are you doing the AWS thing?

kw said...

I fired up the new image with Planning. When I try and launch Workspace I get: Message from the ISAPI plugin:
No backend server available for connection: timed out after 10 seconds or idempotent set to OFF.

Any ideas?

kw said...

Wow... This new Image John uploaded is half the size (25GB) of the previous one... Wonder how? Thats good news though when using Dropbox to get move to my laptop.

kw said...

Well converting it to a VMWare Image works. I just don't have enough RAM on my local machine to run it. John recommends 16GB Ram to run it and I just don't have it. Still pretty cool though.....

Anonymous said...

Sam,

>>I just don't have enough RAM on my local machine to run it.
^^^And that's what drove me to the cloud, and until (and probably even after) I buy a 16 GB laptop, that's where I'll stay.

The offer is there -- if you want to document your AMI->VMWare conversion and blog it here you will get full credit, link to your LI profile, whatever. I am more than happy to have someone else write the posts. :)

Or I'll figure out how to do it. Or maybe even the other way (VMWare->AMI) if it's doable.

Regards,

Cameron Lackpour

kw said...

CL,

It's a dilemna as the charges can add up and when you terminate a image so you don't get charged anymore, all data is lost. That's why I think running it on a image is better. How much is the cloud when it's in a 'stopped' state? Also found some 32GB laptops. Which would really give you enough RAM. Assuming someone could upload one that has 24GB or something. These laptops are $1400 at http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=370579415040

Still contemplating as I just need a sandbox to learn.

Thanks,

Sam

John A. Booth said...

Just using Amazon S3 for the first time to bring down a 100 GB image and so far it's working much better than any alternative I've tried to date. The more I use the Amazon tool sets the more useful I find the different pieces.

Bruno Andrade said...

were you able to get around the workspace error message?
I'm getting the same here and I see that a bunch of Hyperion services are not starting up.

kw said...

Bruno,

What kind or RAM you using. I booted it up with a 4GB RAM and it took forever to fire up. John said you need atleast 16GB RAM machine to run this. I bet if you wait long enough it will fire up. Also, try a 'reboot' within the instance. Eventually it should fire up.

Anonymous said...

Guys,

This is why I went to the cloud. It doesn't work in 8 GB, let alone 4. The compact deployment *might*, but that's not this image.

Sam, FYI, at $0.62/hour, AWS is pretty darn cheap. Short of buying a big laptop, the cloud is a cheap, effective way to work.

Regards,

Cameron Lackpour

Bruno Andrade said...

I tried the 16GB of memory and still getting the workspace error message:

Message from the ISAPI plugin:


Were you able to get around this?

John A. Booth said...

Bruno, look at task manager and ensure the processors of settled down.

Next review the C:\Oracle\Middleware\user_projects\epmsystem1\diagnostics\logs\services and look at the specific services such as workspace and planning and confirm the last few lines have "RUNNING" in them. If not it is still starting up.

John A. Booth said...

Bruno, look at task manager and ensure the processors of settled down.

Next review the C:\Oracle\Middleware\user_projects\epmsystem1\diagnostics\logs\services and look at the specific services such as workspace and planning and confirm the last few lines have "RUNNING" in them. If not it is still starting up.

kw said...

Did you try and 'reboot' the instance?

kw said...

Did you try and reboot the instance within AWS?

kw said...

Yes. On one of the instances I had to manually start the Workspace service, I think... Can't remember exact name. I changed it to 'Automatic' for future.

Bruno Andrade said...

I checked the services and the one for Reporting Print Service was not comming up for some reason, tried to start the service but gives me an error message

other than that, all the others are showing up as running.

Waited for long and still getting the same error message

Dhiraj said...

hi,

I am trying to search on the AMI and it doesn't show the ami (both the one that has been checked and the other one which was written after it). Can you please help ?

Thanks,
Dhiraj

Cameron Lackpour said...

Dhiraj,

John seems to have gotten rid of that one. Try this one instead: ami-09b26360.

It has Planning as well -- so much more fun and I just (like five minutes ago) posted on it as well.

Regards,

Cameron Lackpour

Dhiraj said...

hi Cameron,


Thanks for the quick update. I wasn't expecting such a fast response on my previous comment.

This is my first foray into HFM as my company is planning to implement HFM. Just took the class from oracle over the last week and wanted to get my hands wet and at present we dont have an inhouse instance ready yet, so was searching across the net for how to get an instance.

As such, the installation guide was quite tricky and I didnt had the kind of RAM it needed. So stumbled upon yours / johns blog.

You guys are doing amazing work.

Pretty excited to kick the tires and see how things pan out in HFM.

Regards,
Dhiraj

Cameron Lackpour said...

>>Thanks for the quick update. I wasn't expecting such a fast response on my previous comment.
^^^It's all part of the service. :)

>>You guys are doing amazing work.
^^^That's John, really. I'm just along for the ride.

AWS really is a very useful tool -- the book chapter I wrote was all done in the cloud. It just worked. I really can't ask for much more.

Regards,

Cameron Lackpour

Dhiraj said...

First ride on cloud, so far so good. Used the. Spot bidding and came to like 20 cents an hour.

Pretty darn cheap. Was able to start eom architect right away. No issues at all. Thanks

Here are some questions on aws.

1 when you stop the instance seems like all current data gets wiped off ? So you can't store the application, data at any time or am I missing something ? Like how do I patch the instance / Continue on application.
Charge for standby are also not clear. Any tips how you guys are working ?

2 how to use block storage so that data could be stored ?


Any help on this end will. Definitely set me up for using aws.


Regards.

Cameron Lackpour said...

Dhiraj,

I've never used spot pricing for AWS. 20 cents an hour is cheap. Is that for an m2.xlarge instance?

>>when you stop the instance seems like all current data gets wiped off ? So you can't store the application, data at any time or am I missing something ?
^^^That sounds like you're terminating the instance.

I do shutdowns (also known as stops wtihin the AWS console) from within Windows and leave that stopped server in AWS. Yes, I am paying for the drive space, but it's $0.10/gigabyte/month.

If you terminate an instance, the contents of the drive are *gone*, forever.

If you terminate an instance, go take a look in your volumes -- unless John has changed the way he set up the AMI, volume doesn't go away and you get charged. I have a blog post in the making that will show how to get around that. It is kind of a pain.

As for block storage -- I thought you were looking at HFM -- no block storage there. Or are you talking about how to add additional drive space.

Regards,

Cameron Lackpour

Dhiraj said...

hi Cameron,

Just wanted to update you on my usage on AWS. The On Demand instance is pretty cool. It gives you the ability to use unused capacity of AWS. However, once the instance is terminated, then I am not sure how to now create a new instance with the EBS that is sitting in the user profile.

At a very high level, I could see that it is the Harddrive available and I am getting charged for it (0.10$ / Gb / Month for the storage). I still need to figure two things.

[1] I see you have put a new AMI after the one that you suggested 2 weeks back. I am not sure how to use the new AMI along with the storage that is already provisioned.

[2] How do I use the EBS already created with a new instance, considering that my instance is gone as soon as I terminate it.

Some other questions will also come to mind as I look at this.

Regards,
Dhiraj

Chad said...

Hi Cameron, This is great although I can't see that the configuration for the web client is set up, specifically HFM???

Is this primarily for planning? I am using ami-3173a358

Thanks, Chad

Cameron Lackpour said...

Dhiraj,

I can't comment on how On Demand/spot instances work as I've never used them.

Re the volume hanging around -- I think I've commented about this before -- the way John's AMI is set up means that the volume sticks around when the instance is terminated. You have to manually delete it. I have a blog post (a really long one that I need to get out) that shows how to use the AWS API from the command line and change the AMI (you need your own copy but that's easy to do) so that the volume gets deleted on termination. You *can* attach the volume to a new AMI if you need anything off of it.

If you kept John's old AMI running (as I have as I used it to write my book chapter and I am *not* going through the pain of installing ODI 11.1.1.5 unless I must) you could STOP it, and create your own AMI. I have done this.

You need (unfortunately) to be cognizant of AWS' EC2 three drive states:
1) Stopped -- A server instance is stopped and can be restarted at any time.
2) Running -- A server instance is running.
3) Terminated -- A server instance is shut down and destroyed.

In the case of #3 as I wrote above, there is no way to use that volume to start back up. (Okay, if you have a bigger brain than me, you can reattach detached volumes, I think, but that's beyond what I can intelligently talk about. I'm sure you can find something like that on the web).

I hope that helps and I'm glad that you're finding it useful. I *love* AWS. Amazon ought to be paying *me* for my proselytizing. I guess they really ought to be paying John. :)

Regards,

Cameron Lackpour

Cameron Lackpour said...

Chad,

This is *not* an HFM AMI as you noted, but Essbase and Planning.

However, you can now install HFM onto the AMI for your own testing use.

Btw, I have been doing some testing in Planning and I will state that I have some pretty lousy performance on a M2XLarge instance with OutlineLoad.cmd. I really should try it with a big, big, big box to see what the performance is like. One day...

Regards,

Cameron Lackpour

Dhiraj said...

Cameron,

One question on start and stop script. Is there a standard script to start and stop all services which is available ?

I have installed Hyperion on a server and need the correct procedure to put all the services that need to be started and stopped in a script.

Appreciate your help in this regards.

Anonymous said...

Is there a password for Workspace or Essbase? I can't seem to login into Hyperion...getting invalid login error.

Cameron Lackpour said...

Anon,

I'm not sure if the Essbase-only AMI is still available, but if you found John's Planning-enabled AMI, you should be good.

In any case, the password for both AMIs is "epmtestdrive".

Regards,

Cameron Lackpour

Anonymous said...

I am a complete cloud computing novice so please spare me for asking this. AWS includes free usage with new registrations. Those aren't enough for running the Planning instances?

Vikas said...

I got it to work but whenever I click the workspace link I get this error:
"No backend server available for connection: timed out after 10 seconds or idempotent set to OFF.
"

Any ideas?

Cameron Lackpour said...

Anon,

You need to use an m2xlarge instance to run any of this stuff. The free instances are just too small.

If you want everything I know (pretty much) about AWS/cloud go to www.odtug.com and download my whitepaper and the presentation I did last year with John Booth. It's all there.

Regards,

Cameron Lackpour

P.S. In the cloud right now working on my KScope12 presentation. Ugh, what else would I do on a Saturday morning? You should come to the conference.

Anonymous said...

Cameron
I realized that after reading the server specs. Micro instance with 600+ MB ram. Too low.
I was able to start the instance as per your walkthrough.

Though I seem to be running into similar error as Vikas - "No backend server available for connection: timed out after 10 seconds or idempotent set to OFF."

I found a walkthrough on how to actually install 11.1.1.3 on EC2. It might be a way actually install 11.1.2.1 individually:
http://hyperion-talk-n-menu.blogspot.in/2009/11/oracle-hyperion-epm-bi-v11113-on-amazon_08.html

Downloading your whitepapers now. :)

Cameron Lackpour said...

Anon,

If you want to install 11.1.2.whatever I wish you good luck. It's certainly doable. The reason I use these cloud images is so I *don't* have to do an install but to each his own.

Re your error -- my guess is that the services haven't started up completely.

Hmm, looks like Apache isn't running?
https://forums.oracle.com/forums/thread.jspa?threadID=1091791

Regards,

Cameron Lackpour

Anonymous said...

Hi all the I have followed all the steps outlined but cannot locate the AMI machine with the EPM software installed.

Can anyone help with this?

Cameron Lackpour said...

The AMI in this post has been superseded by one that includes Planning.

Take a look here: http://camerons-blog-for-essbase-hackers.blogspot.com/2012/02/copy-of-two-no-three-tips-from-oracle.html

The new AMI is: ami-3173a358

Regards,

Cameron Lackpour

Yansane Yansane said...
This comment has been removed by the author.
Cameron Lackpour said...

Yansane,

The AMI is there (I just checked). Are you searching in the Region us-east-1?

Regards,

Cameron Lackpour

Yansane Yansane said...
This comment has been removed by the author.
Cameron Lackpour said...

Yansane,

Use as an id "Administrator". Your local PC's id will not work.

Regards,

Cameron Lackpour

Franklin Twain said...

I was able to log onto and connect to the instance. However, when I attempted to open the "Workspace shortcut," I received the following error

"Error 404--Not Found
From RFC 2068 Hypertext Transfer Protocol -- HTTP/1.1:
10.4.5 404 Not Found
The server has not found anything matching the Request-URI. No indication is given of whether the condition is temporary or permanent."

Please forgive my ignorance. I am trying to tech myself Essbase and I am getting off to a slow start. Any help would be very much appreciated.

Thank you,

Thomas Rausch

Anonymous said...

Hello Cameron,
another question - when the original AMI image is removed, lets say it is replaced by a newer version, then all copies stay as they are until the copy is terminated? So there is no link between the AMI I start and work on and the instance John had originally created?

Thanks in advance for your reply.
I hope to meet you at Kscope13 in New Orleans!

Regards from Holland.
Philip.

Unknown said...

What happened to all the screenshots?

Cameron Lackpour said...

Chris,

Blogspot sometimes eats pictures. It's weird, it's undocumented, but if you look around, it happens to lots of people.

I need to go back and fix all of these posts (older ones, mostly) but I struggle to find the time. Also, it's really annoying.

The one thing I can say is that 11.1.2.1 is a while ago -- I am surprised you are still reading about it.

Regards,

Cameron Lackpour

Anonymous said...

Hi Cameron,
What is the latest AMI to use for Essbase

Anonymous said...

Hi Cameron,
i am using ami-3173a358 with m3.large instance($0.266 per On Demand Windows m3.large Instance Hour).Can i use $0.133 per On Demand Windows m3.medium Instance Hour.
Regards
Raj

Anonymous said...

Hi Cameron and John, Great AMI. I was able to connect using remote desktop. Login to EAS, created cube. Now, I try to use Smartview from my local machine. When I logged into workspace my URL was: http://metavero/workspace/index.jsp And not http://www.metavero.com:19000/workspace
Could you please advise on how can I get SmartView working from my local machine? I tried both shared and private connection with or without 19000 in the URL, with no success. Thank you.

Anonymous said...

Hi Cameron,

Is this no longer publicly available? I am checking John's blog too just in case it is just a matter of a newer AMI id

But thanks anyway for the introduction to AWS, excellent tutorial as always and if all else fails this and related blogs will definitely help me a lot if I decide to instal my own.

Cheers
Y

Cameron Lackpour said...

Y,

Sorry, the AMI was pulled long ago.

Fwiw, I've gone to a Really Big Laptop and just installed it myself.

Jake Turrell has a good series on how to do it yourself. I did it on my own and hated every minute of it but in the end got it to work.

Regards,

Cameron Lackpour