Hi,
really its a easy & quick intallation.

Would be easy use the google big-table ?

Hi Paco,

Google's BigTable isn't a relational database, so it's not possible to use it to store JIRA and Confluence's data.

However, this isn't a problem: If you follow the instructions above, your JIRA and Confluence server will be automatically configured with a PostgreSQL server to store all your data.

Very cool Adrian!

Excellent. Can't wait to hear how to use a domain name so it is easily accessible.

You've done an awesome job with this Adrian! Respect!

This is great!!

anyone got any pointers on the stable domain name?

Hi danb,

I've just followed up with a post on setting up a stable domain name.

Does this work if I have a server on SliceHost?

Hi Paul,

At the moment, this tool only supports Amazon EC2.

Does it matter how much CPU usage the app is using, - or simply that it is 'active' therefor it is being charged?

$16.80 is not ideal if its just sitting there idle, - for sure!

Hi James,

The charging is independent of how much CPU is being used. If a server is running, but not being used, you will still be charged at the full rate.

There are two ways to counter this:

1. Pause the server when it's not being used, using the instructions I give above.

2. You may be able to make significant savings by making an up-front lump sum payment to Amazon. I'll cover this in a later post.

I got my confluence to use the JIRA database for authentication... I think that would be a great addition to this setup.

Thanks danb,

That's a great idea.

If there's an update to the image, I'll definitely look at implementing it.

This is good,

I recently went set up crowd, jira, and confluence on an ec2 instance. I found it was more involved than i expected to integrate the three. Any thought on adding a pre-cooked image adding crowd authentication to the mix?

Hi Jonathan,

It's a great idea, but it's probably best that this image just contains JIRA and Confluence, as it was developed for small teams who take up this week's $5 license offer.

Crowd is a great solution for larger organisations looking to implement single-sign on, but it's probably overkill for small teams.

Maybe this is something we'll consider in the future.

I've been asked if this configuration supports JIRA and Confluence plug-ins.

The answer is, no, not as yet. We might add support for this at a later date.

I assume that instead of the "Your_EBS_Volume_ID" you meant the "Instance ID" for the server?

Thanks for this - great to have Confluence/Jira up and running in minutes - I love Confluence and have been meaning to test out JIRA for some task/request/issue tracking at my company. I've kept looking for an affordable wiki that I liked for my personal use, but I kept coming back to Confluence after getting fed up with the others.

Hi William,

The "pause" command will accept either an EC2 instance ID or an EBS volume ID (which the tool will automatically map to the attached instance ID), but it's probably simplest to use the latter, as this won't change each time you pause your server.

All other commands accept an EBS volume ID only.

Any chance Atlassian will put together a similar integrated installer for the Jira + Confluence combo that will work on an Ubuntu server or a Windows server? Since we are limited to "standalone" versions, it would be nice to only have a single Java server instance runnning for the combo.

Hi Adam,

I'm not aware of any current plans to do so, but it sounds like a great idea to me. I'll pass it on, and we'll see what happens.

By the way, you can still use your Starter license with the WAR editions of JIRA and Confluence. The only restriction is that we're not able to provide support for this configuration. We recommend that you stay with the much simpler Standalone editions.

Plug-in support, please!!!!

Hi Sean,

Plug-in support is definitely something that I'll be looking into.

It would be really helpful to hear which plug-ins you'd like to install. Check out the plug-ins that are available for JIRA and Confluence.

Keep your ideas coming. You can add them to the JIRA project I've created to collect feature requests. I've already added your request for plug-in support.

Many kudos to select Postgres for this instant-on (almost) solution!

Where is the RSA key that the jar uses? I forgot to copy it when it was first displayed the first time the program ran.

Hi Randy,

There's no way to retrieve the existing private key after it is generated.

You can easily generate a new key pair by deleting the existing "instant-atlassian" key pair (I recommend you use the AWS Management Console for this task). The instant-atlassian tool will generate a new key pair next time you run it.

How do I ssh into this instant instance? Would I have to bundle the instance on EC2 and reissue our own keypair file?

These are the plugins we use all the time:
* Scaffolding plugin
* Reporting plugin
* Linking plugin
* Replace and render
* Table plugin
* Survey plugin
* Content formatting macros (especially {table} and related macros)
* Advanced Search plugin
* Label-Tools plugin

Hello,

Great post!
May I ask what happens behind scene? e.g how to do the same thing through https://console.aws.amazon.com

Thanks in advance

Hi Michael,

No, you don't need to rebundle the instance to connect using SSH. Just follow these steps:

Take a note of the "private key" that the tool gives you the first time you use it (If you missed it, delete the "instant-atlassian" key pair from your AWS account, and run the tool again). Use this key pair with these instructions. Skip the instructions on "authorising access", as the tool has already taken care of this for you. Just be aware that any changes you make to the boot disk will be lost when you "pause" your server, as this disk is not persistent.

Hi Eriks,

It's not possible to reproduce everything above using the AWS console, as the instant-atlassian tool passes the instance some metadata that the instance requires to attach and format an EBS volume, but here's a summary of what's going on behind the scenes:

create:
An AWS security group and key pair (each named "instant-atlassian") are created if they don't already exist.
An EBS volume is created.
An AWS instance is created from the Instant Atlassian AMI.
The EBS volume is attached, formatted and mounted in the instance.

pause:
The AWS instance is terminated.

resume:
An AWS instance is created from the Instant Atlassian AMI.
The EBS volume is attached and mounted in the instance.

delete:
The AWS instance is terminated (if one is running).
The EBS volume is deleted.

Do the scripts assume that I'd like a US based volume. I'm in Europe and I'd like to use this in the eu-west region.

Hi Ian,

Yes, the tool is currently hard-coded to locate your server and storage in the US, rather than Europe.

If you're concerned about network latency, I've been using the tool from Australia, and haven't found latency between here and the US to be a problem.

The choice of European-hosted servers is something I'll consider for any future release.

From the UK, the ping time is around 120msec, which is quite slow. I'll keep watching.

Thanks, making progress...

IanM

The $5 deal was great. We're now testing it out here at work. I heard people using it, but it always seem to be too expensive. Now at least we can try, and using EC2, it's a piece of cake! Thank you. GM

Hi -
I created a how-to hook it up to DynDNS: http://www.ausbusiness.net/review/confluence-amazonec2-dyndns/ and how to create start up batch scripts for windows: http://www.ausbusiness.net/review/start-stop-scripts-confluence-jira-amazon-ec2/

Hope it helps.

Guennadi M

Is it possible to upgrade the $5 versions to the new version 3.0, or is it back to the normal licence costs?

Could I just confirm that it is possible to use the jiraissues macro within confluence in this configuration. I can see the XML rss feed but am getting an "APP_UNKNOWN; Unknown Application: {0}; ["confluence:2718366"]" error when I save the page.

The url is something like,
http://ec2-67-202-13-83.compute-1.amazonaws.com/jira/sr/jira.issueviews:searchrequest-xml/temp/SearchRequest.xml?pid=10000&resolution=-1&component=10000&sorter/field=priority&sorter/order=ASC&sorter/field=summary&sorter/order=DESC&tempMax=1000

Getting on really well with Confluence tho'

Thanks

IanM

Hi Ian,

Your $5 dollar starter license entitles you to a year's worth of upgrades, so, yes, you can use it to run Confluence 3.0, released last week.

At the end of each year, you will have the option to renew your support and maintenance for another $5, which will also be donated to charity. You can find answers to other common questions about the starter license in our FAQ.

I'm working on some improvements to the Instant Atlassian tool and image now, including an upgrade to Confluence 3.0. I hope to make these available in the near future.

adrian ... fantastic tool! i've been using it with success for the pas few weeks. the last couple days, though, when i issue a pause command, i get this:

Terminating EC2 instance i-xxxxxxxx... java.lang.Exception: Unrecognised state for EC2 instance i-xxxxxxxx: running

did aws change some response and break the tool?

thanks!

Hi Michael,

Thanks for letting me know.

I can see the problem, and will have a fix out shortly. Your attempt to pause the instance is probably succeeding, despite the error.

Until the fix is available, you can use the AWS Console to check that your instance is being terminated properly when you use the "pause" command:

http://aws.amazon.com/console/

Cheers,
Adrian

Hi Adrian,

I had missed the boat on the $5 JIRA license promo. I am currently looking at running JIRA on AWS. I suppose I need to get a regular license? If so does your instructions differ (Steps 1-5)? Also curious, in Step 4, assuming the instance tool is still applicable, you wrote "some of the options have been pre-configured for you, and are disabled in this configuration" - what has been disabled and which edition (Standard, Profession, Enterprise) of JIRA is packaged in the instance?

Many thanks
Lisa

Hi Lisa,

You can use any valid JIRA Enterprise and Confluence license with the steps above, including the free personal licenses for JIRA and Confluence.

The options that have been pre-configured are the database connection details, and the location of data directories on the server.

The image includes the Enterprise edition of JIRA.

Adrian

I did this back in April. Worked great! Thanks. It would be fantastic to be able to do it again now and have JIRA 4 and Confluence 3 install from the image. Any plans to do this please?

Hi Alek,

Thanks for your interest. This is definitely something I would like to do.

Now that JIRA 4 has been released, I will be taking some 20% time to update Instant Atlassian with JIRA 4 and Confluence 3.

Until I have this ready, you might like to follow this guide we've developed for those who'd like to set up and integrate all of the products included in our $10 starter license offer.

We just took advantage of the $10 starter license offer and plan on deploying to AWS. We would like to use "Instant Atlassian" to create our server, but it sounds like you are still working on that. Do you have an ETA on when the new AMI might be available?

Thanks!

Hi Thad,

I'm sorry, I can't provide an ETA for this, as it's not part of our scheduled development.

If you need a hosted solution, you should consider our hosted products, JIRA Studio, JIRA Hosted and Confluence Hosted.

JIRA Studio combines JIRA, Confluence, FishEye, Crucible and Subversion into a single, integrated product.

The benefit of our hosted products is that we look after backups, upgrades and other maintenance for you, and we provide full support should you need it.

Hi Adrian,

I bought $10 starter licenses and complete the Atlassian Dragon Quest at Amazon Cloud with EC2 and S3( for all data and configurations files). After that, I rebundle my AMI.

The problem is that the key for the atlassian products must be provide every start of the instance.

I would like to take advantage of start/stop scheme like you show at this post and then automate all process.

Can you have any tip for solve the problem of provide all license (six) for each restart ?

Thks!!

Hi Adrian,

Sorry, disregard the problem about provide the keys. I create a new image and rebundle them. Now it works!!

Tks!!

ZoneInfo: /System/Library/Frameworks/JavaVM.framework/Versions/1.5.0/Home/lib/zi/ZoneInfoMappings (No such file or directory)
ZoneInfo: /System/Library/Frameworks/JavaVM.framework/Versions/1.5.0/Home/lib/zi/ZoneInfoMappings (No such file or directory)
ZoneInfo: /System/Library/Frameworks/JavaVM.framework/Versions/1.5.0/Home/lib/zi/ZoneInfoMappings (No such file or directory)
ZoneInfo: /System/Library/Frameworks/JavaVM.framework/Versions/1.5.0/Home/lib/zi/ZoneInfoMappings (No such file or directory)
ZoneInfo: /System/Library/Frameworks/JavaVM.framework/Versions/1.5.0/Home/lib/zi/ZoneInfoMappings (No such file or directory)
java.lang.NullPointerException

the jar tool was run on Mac OS X 10.5 - software update says there's nothing to update, but I don't exactly trust Apple's handling of Java (like how to get JDK)

java -version
java version "1.5.0_20"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_20-b02-315)
Java HotSpot(TM) Client VM (build 1.5.0_20-141, mixed mode, sharing)

any thoughts on what's going wrong?

Hi Joe,

I don't think this problem is anything specifically to do with the Instant Atlassian tool, because it was developed and tested on a Mac.

My colleague Nick has just run into a similar problem with Java on Mac OS X 10.5. We think it might be because he has changed the "CurrentJDK" soft link.

The recommended way to change your default JDK in Mac OS X is to use /Applications/Utilities/Java Preferences, or to set the JAVA_HOME environment variable.

I hope this helps.

Extremely useful effort. A lot of new customers are setting up 10 user installations of the full Atlassian suite right now. It would be fantastic if this tool were updated to install the entire suite, deployed to an EC2 instance. It's too bad getting the suite running is difficult enough to earn people tshirts...

Any chance this could be given some priority? It can only help Atlassian's bottom line, since customers buying expensive licenses would use it too.

Any plans to make this a paid for AMI and skip the licensing requirements? Also, any plans for JIRA 4 (this is 3.13.3).

Hi Adrian, I planning to purchase the licenses but waiting on the new image with the updated version. I wonder if now you would have the ETA? This clearly generates a lot of interest but doesn't seem to get a priority,,,

@Joe, here is a link to a blog I wrote back in July about Java on Mac OS X.

http://www.rich-software.com/wp/2009/07/switching-java-versions-on-a-mac-setting-java_home-and-using-java-6-revisited/

It just might help you understand how to fix your issue.

====

Hi Adrian,

I took advantage of the $5 Confluence and JIRA offer back in April and just got around to installing Confluence on a Mac Pro behind Apache. I also am using DynDNS to make it publicly available to my small team. I have a couple questions, though. Some things are acting strangely for me when running directly from the machine that is running everything. I am accessing it via the dyndns.org domain and I do not get the dynamic search lookup feature and I get a blank search results page no matter what I type. Search works for the remote team members, though.

I suspect my issue has something to do with all the redirection going on. Confluence is running at localhost:18080 (the Server Base URL) and Apache is redirecting nicely to this using SSL on the public side. I could not find anything that helps me configure this such that everything works for me and the remote users. I've tried setting the Server Base URL to the external domain name, but then other things behave badly. FYI, the main features seem to work for me -- e.g. adding/editing pages, administering Confluence, etc. Search is the only one I can't get working - at least that's the only one I aware of at the moment.

Once I get this figured out, I'll post a blog to help others as an even less expensive way to get up and running with your most excellent tools!

Any tips on what might be wrong with my configuration? I could post the Apache and Tomcat configs, which I modeled after what I found on your site.

Thanks so much!
Rich

Hi Rich,

Thanks for the information on the Mac OS X JDK.

You should have your Confluence base URL set to match what your users actually enter in their browser to get to Confluence. If you're using SSL, make sure that it starts with "https:". Also, it needs to include the "context path", which is "/confluence/" in the Instant Atlassian setup.

So, your base URL should be something like: "https://your.host.name/confluence/".

I hope that helps you fix the problem.

Just one more vote for an update to support JIRA 4.

We have already purchased licenses for JIRA 4 and have an existing EC2 infrastructure we'd like to leverage. the 1.1 version is a great tool, but is there no other way to upgrade to JIRA 4?