AWS hosted Atlassian - requirements

[lozog95]

Dear Community,

I have a question regarding hardware requirements for hosting such applications as Crowd, Jira Server, Confluence Server, Bitbucket Server and Bamboo Server on AWS EC2 instances with RDS (postgres). I cannot find it in the documentation. My instances are licensed as following:

• Jira Software - up to 2000 users
• Jira Service Desk - 25 agents
• Confluence - up to 2000 users
• Bitbucket - up to 1000 users
• Bamboo - 10 remote agents

Each of those applications has its own EC2 instance and RDS. I'm wondering what are the recommended db and ec2 types provided by AWS? Is there any article I could read?

Thank you in advance for help.

Best Regards,

Lukasz

[Alexis Robert · Accepted solution]

Hi @lozog95, 

I think the instance sizing will depend not only on the number of agents but also how you use each component.

For example, in Bamboo it's more about the number of plans, their complexity and the size of the projects than the number of users. For Jira you'll need more powerful servers if you have lots of custom fields, post functions, custom scripts.

These pages should offer some guidance on hardware requirements:

• Bamboo
• Confluence
• Bitbucket
• Jira

In my experience, it will depend a lot on the customization of your applications and the number of concurrent users. 

For your smaller instances, a t2.medium/large should be enough. For the bigger ones, I'd look at the c5.xlarge ones.

--Alexis

[Kumar]

Hi @lozog95 Have you resolved this if you resolved this can you please suggest me the steps 

I have similar requirement I have a Jira Server in AWS Ec2 instance and postgres db also in the same instance 

Here I need to Migrate the Postgres DB of jira which is in AWS Ec2 instance  to AWS RDS db 

Can you please suggest me how to achieved this 

Thanks,

Kumar

[Kavirajacholan Selvaraj]

Hi @Kumar 

Do you resolved this? Any idea what is the capacity you have used here and completed the migration?

[Kumar]

@Kavirajacholan Selvaraj  I resolved this i have sucessfully migrated the postgres DB  to AWS RDS postgres

Here are the steps to follow:

1. Login to Jira  server application  Database by using the below command
   
   psql -U postgres   
     or 
   su - pgsql / postgres
   
2. Create the " dump " file of the application db by using the Below command. 
      pg_dump  (dbname)  >  (dbname.dump)
   
   
3. It Will store the dump file in  this location " /var/lib/postgres/  ".
4. After that try toexport the dump file to AWS RDS DB  by running the below command.
   
    " psql -h <endpoint> -U <username> -e <dbname>  <  <dumpfile name> "
   
   // endpoint : RDS instance "Host name / endpoint ".
   // username :  RDS DB user name
   // dbnmae : RDS created dbname
   // dumpfilename:  that you created a dumpfile that application which is using.
   
   the original command looks like this "psql -h jiracc.xxx.xxx.amazonaws.com -U "RDS User name" -e "rdsdbname" < db.dump
   
   after run this command it will Export the all the DB data to RDS db. 
   
   
5. Before run the above command you have to create a "Database and user"  and that user must be a superuser in that "AWS RDS instance".
6. After Exporting Modify the Jira application "dbconfig.xml" file.
   
   Before Modifying the dbconfig.xml file
   
        <url>jdbc:postgresql:5432/jira</url> 
   

   and

         <username>username</username> -→ DB username

   and 

         <password>XYZ</password> -→ password 

   
   
   After Modifying the dbconfig.xml file
7.     <url>jdbc:postgresql://jira.XXX.xus-amazonaws.com:5432/jirad</url>
   

   and

        <username>username</username> -→ DB username

   and 

         <password>XYZ</password> -→ password

8. After Modifying Stop/Start the  Jira application.

[Kavirajacholan Selvaraj]

Thank you @Kumar While migrating Confluence do you see any 14+ hours to restore the backup?

We are targetting datacenter server to AWS. Jira is on Windows with MSSQL. Now I am migrating it to Linux with Postgresql.

Jira migration is sucessfull in 45 minutes. But Confluence taking 14+ hours after increasing Heap size its 7 hours.

Any idea how to overcome from this timeline?

[Kumar]

Hi @Kavirajacholan Selvaraj  sorry for late reply 

I have migrated Jira only we are not using confluence  and we are using jira Server.

Thanks,

Kumar

[Kavirajacholan Selvaraj]

Sure np. thank you!

[Kavirajacholan Selvaraj]

@Kumar Have you used EFS for attachments while migrating to Jira? When I did xml backup import in new server.

I changed the BaseURL

Now I am not able to configure Attachments, Application links, Avatars, Logs. Any idea how you achieved that?

Please advise.

[Kumar]

Hi @Kavirajacholan Selvaraj  i have not used any EFS I directly Dumped the Old db into New RDS db 

I have followed the AWS documentation

https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/PostgreSQL.Procedural.Importing.html

Thanks,

Kumar

[Kavirajacholan Selvaraj]

Thank you @Kumar Do you configured SSO and faced any issues like this? 

"You must set a valid base URL."