Release Reference Application modules for 2.4 release

I’m not able to pull up nor SSH into that box. Are we waiting on help desk support? Or are is anyone else able to SSH into the machine?

This is the last error message that @jdegraft reported to me on IRC before i told him to try helpdesk:

Ah. Looks like the same problem I’m seeing: unable to SSH into that VM.

@dkayiwa, @jdegraft – please confirm that a help desk ticket has been created, they understand the priority, and check in on status daily. Thanks! If you have not heard anything from the help desk within 24 hours, then please let us know here.

A help desk ticket has been created last week. I am still waiting for the fix.

@jdegraft, please reply to that help desk ticket (CC’ing myself & @dkayiwa) stating this is a blocker for Reference Application 2.4 release, any help on restoring SSH access to uat02 would be appreciated (including completely resetting the VM). If the help desk is blocked – i.e., they are stuck waiting on the VM host to take action – then we can ask them to elevate that ticket and, in the meantime, look elsewhere for a temporary solution (e.g., I could pay for a digital ocean VM and have the help desk point to it until uat02 can be fixed).

Help desk says uat02 is not meant to support ssh.

It is not clear that uat02 can be setup for CI deployment. Some guidance on how to deploy ref app 2.4 to uat02 is needed.

This is still a blocker on ref app 2.4 release to UAT.

Hey there all – please keep in mind that the help desk is not paid and are entirely volunteer – so please keep in that in mind and set your expectations accordingly. You won’t get immediate action – sorry about that. Beyond this, I do not know. According to a document I found, I don’t know if resources have been assigned for uat02. This link requires that you use your OpenMRS email address.

@michael or @ryan can comment farther.

Thanks @r0bby. I was confused because we’ve had a uat02 in the past (must have been reclaimed) and both and respond to ping requests.

I’ve submitted a new help desk request (#27784) asking that devtest05 be reclaimed (re-assigned from @surangak to @maurya) to be used for uat.

What we need most is a target for UAT for the reference application. So, in decreasing order of importance:

  1. SSH + sudo access for @jdegraft and our CI server.
  2. Point to the machine.
  3. SSH + sudo access for others to help @jdegraft.
  4. Rename devtest05 → uat02.

#1 is the what we could use most.

I am not sure it will get done immediately. I have to pack for OSCON and would rather not be doing this while in Austin.

Ack! It looks like I’ve got two requests in to the help desk that conflict:

  • Need QA for platform [#27438] on May 2nd, asking to convert devtest05 to qa02 (aka qa-platform)

  • UAT for reference application [#27784] on May 12th, asking convert devtest05 to uat02 (aka uat-refapp)

@raff, per this page, you are using two devtest servers. Which of these would you like to give up to be recommissioned as qa-platform?

I’ll send you and @surangak my public key, so you can add an account for me as sudoer and we can move forward while the help desk volunteers are busy with other things.

We’ll use devtest05 as uat-refapp and whichever of these two you need least for qa-platform (I’ll update help desk ticket #27438 accordingly).

Actually, @helpdesk, can we please re-title some of the assignments listed here so they are easier to understand? (GW89) - Rafal Korytkowski (Metadata Sharing Builder)

Assuming this is the server, this should be assigned to “Reference Application Team” and its title should be “Metadata Builder for Reference Application” (GW152) - Maurya Kummamuru (OpenMRS Reference Application 2.2 Testing)

This should be unassigned. Reference Application 2.2 was released in April 2015 – this was the UAT server we used for the release candidate, and it hasn’t been used for this purpose since April 2015.

@burke, to be explicit, assuming that the helpdesk assignments article we’re referring to is up-to-date, (GW152) should be free to commandeer, and we don’t need to touch devtest00 or devtest04 (for this purpose).

I’m just going to be explicit – a large chunk of the infra team is going to be in Austin for OSCON, and I believe @ryan is away. @maany is busy and perhaps recovering from exams. So unless a server goes down, ro something like that, this is going to have to wait. Sorry. If you want this to not fall through the cracks, ping one of the open cases for this issue.

I confirm that I no longer need (GW152) to be assigned to me

I just add that I’d prefer to keep devtest04 for demoing Android Client and OWAs. We are using it actively.

@darius, @michael has been trying to get the testing server page updated & maintained; however, It’s not quite up-to-date, especially wrt purposes/assignments. I’m trying to get us to our target of having these six dedicated machines for our CI process.


  • Automated build, automated testing for validating commits
    • int-platform, int-refapp
  • qa (automated build, manual testing) for testing new & in progress features
    • qa-platform, qa-refapp
  • uat (manual build, manual testing) for testing releases
    • uat-platform, uat-refapp

Current status and TODOs

Unfortunately, we don’t have the ability to add new machines right now. I thought we already had a uat-refapp ( server since it was responding to pings with the same IP address for our uat-refapp. But, apparently, this is an artifact and the machine these DNS entries are pointing to isn’t actually available to us. That means that devtest05 is going to become uat-platform (

That means that we need to decide if these devtest machines are higher priority than having a QA for the reference application:

@raff says he is actively using devtest04. @darius says we need to keep devtest00.

@darius, could you give up devtest03 for qa-refapp?

(The Help Desk has asked for us to appoint a “QA lead” to help coordinate these requests, rather than fielding a bunch of sometimes conflicting requests from different people. For now, given lack of an identified QA Lead for the community, I told them I would backfill this role, so let’s figure out what we want for our CI machines and I’ll take care of summarizing for the Help Desk)


Just a reference to a recent post I made in case you didn’t see it:

Unfortunately, it probably just makes things more complicated… right now, most new commits don’t show up on the int server, because it is mainly using non-snapshot versions of modules.


Can we include all the translations that are done in transifex for the modules, as part of the module released in RefApp 2.4?

I guess at this point we can free up the devtest03 server, since there is no longer active development on the Ebola system and we are no longer actively demoing it.

Give me some time (a week?) to make sure I can replicate the setup on digital ocean, and retrieve any good demo patients, and take some last screenshots.

-Darius (by phone)

Basically it’s too late for RefApp 2.4. (The ~30 modules were actually released back in March, without pulling in transifex translations.)

See this comment: