[Ops] Stale cloudfront cache for Influxdata repo?

Site Feedback
1

I just tried to set up a couple of VMs in Azure’s East US 2 region, and attempting to install telegraf failed with an HTTP 404. The version being installed by apt-get is 1.5.0-1, while the currently available packages on the Debian (wheezy/stable) repo are 1.5.1-1.

After a bit of troubleshooting, I came to the conclusion that the content being served by CloudFront from the cache seems to be stale in some parts (or, the changes have not yet propagated). When I attempt to fetch the Packages.gz file used by the aptitude system/programs, the headers show different values for Last-Modified (8 Jan v/s 11 Jan). Sample output is available at this gist.

I confirmed that this is not a transient issue, by trying apt-get commands from my provisioning tool, from the CLI, as well as manually fetching the Packages.gz (as seen in the Github Gist above). I can also reproduce this on multiple VMs in the same region/DC, but this issue does not reproduce on other Azure regions.

Is this the right place to post an issue like this? If not, kindly redirect me to the appropriate location/forum. If you need additional information on this, please do let me know. I’ll try my best to fetch more debug logs.

2

Sorry about the problem and confusion this caused, you are correct about the cause of the issue and I believe this particular issue should be fixed now.

Is this the right place to post an issue like this?

You can report issues like this here or on the telegraf github.

3

Thank you for the quick reply and resolving this issue! I can confirm that I can access the right packages in the East US2 region in Azure.

Apologies for the delay in responding, I didn’t have an opportunity to test this out in a VM in that affected region till today.