Clojurians Log v2

morning team, I’m back on building out my Roll deploy. Hitting a snag due to the fact that I use lein to build my artifact. it builds the jar correctly but it appears that mach “produce” uses stdout from it’s sh cmd. how would I best use lein in this type of “produce”?

scratch that, I RTFM’d and found that upload! is what I needed

progress update: I’ve deployed my app, I see it running in the console. that’s a big step

Q: I had to create an EC2 keypair but not anything in KMS. this seems to contradict the docs

also, forgive my noobness but how to do see logs for the servers? do I ssh into them like the bad old days to do this? what’s the best practice for this? I would presume OS logging to Reimann but interested to hear what works best

just tested a new artifact deploy. it correctly starts new ASG and destroys the old one. target group also updated. so, in general, it looks good

@steveb8n I think you can look at Amazon CloudWatch.

thx Dom, that does look nifty. I also saw that reimann-tools can do most of what I need so I’ll play around with that first I think

yeah, we use riemann for our logging. But when there's a problem we ssh onto the machine.

ok, next question. my instances come up fine but I’m struggling to get the docker container to build/run my zipped file. I had to use a different AMI id from the one in edge because that one was not available for some reason. Is that the likely cause?

yeah, that AMI is probably one on our AWS account: cc/ @jonpither bug report

The AMI I am using is failing to start the apt-get script…

+ apt-get update Reading package lists... E: Could not get lock /var/lib/apt/lists/lock - open (11: Resource temporarily unavailable) E: Unable to lock directory /var/lib/apt/lists/

can you recommend another public AMI that is similar to the one used in edge?

I notice it is not using docker containers and I would like to use docker with roll deploys in future but happy to avoid it for now

We use an arch linux ami: https://www.uplinklabs.net/projects/arch-linux-on-ec2/

Would require you getting to grips with pacman to run things (install is pacman -S pkgname).

cool. I’ll see if I can find something link that on AWS

it might just work….(hope)

trying now

FYI I am having to manually disconnect some AWS components to get terraform apply to work.

once I see a pattern I’ll log an issue

I’m deploying to Sydney region, none of the AMI’s listed are available so I searched for a recent arch linux image and tried that but…. apt-get is not available to the root account in the image I tried

shall we wait for @jonpither to decide on an “official” AMI to use with roll? probably better than guessing

hmm, arch linux doesn’t seem to have apt-get (as you say, pacman instead) so I think the roll AMI is not arch. I’ll wait for @jonpither on this one

What on earth has @jonpither been up to 😆

oh, I see. Hold on.

@steveb8n which ami did you use? ("RZr9WWrHIS"?)

So apparently we have some base ami, that appears out of nowhere, called "juxt-blue"...

I used ami-1e455f7d

arch-linux-lts-hvm-2017.08.24.x86_64-ebs

I’m gonna try a generic debian since that is where apt-get comes from, last try 🙂

yeah, that sounds like a safe bet. Ignore my arch linux advice 😛

hmm, not many good debians in AU region. I’ll just pick one to test the theory

Okay, well, you apparently just identified a hole in our project documentation. I have no idea where those amis came from... cc/ @jonpither youbrokeit 😛

fyi I have seen a pattern in updates….

aws_launch_configuration.dev_nextdoc_0_0_1_4_g33d9144_dirty: ResourceInUse: Cannot delete launch configuration dev00d8d91630982048de1febebeb because it is attached to AutoScalingGroup dev-nextdoc-0_0_1_4_g33d9144_dirty status code: 400, request id: 9274b4b3-9220-11e7-8faa-03636973240e

I fix by manually deleting the ASG each time

BTW I don’t mind all these little roadblocks. I’m a sado-masochist in that I enjoy devops

I think you can also do something like terraform taint aws_launch_configuration.dev_nextdoc_0_0_1_4_g33d9144_dirty too

oh ok, I’ll read up on ‘taint’ since I’m a TF noob

no luck with random debian…

Processing triggers for libc-bin (2.19-18+deb8u6) ... + apt-get install -y openjdk-8-jdk Reading package lists... Building dependency tree... Reading state information... Package openjdk-8-jdk is not available, but is referred to by another package. This may mean that the package is missing, has been obsoleted, or is only available from another source E: Package ‘openjdk-8-jdk’ has no installation candidate

let’s wait for @jonpither