Cannot deploy because of insufficient memory

Okay thanks. I think I was actually incorrect; looks like my DB is still unhealthy, rather than it being a connection error.

Any idea when I will likely be able to deploy to YYZ again?

Checking into whether there’s an issue with any hosts in yyz.

1 Like

The volume usage display is very new and might show 0 for a while if your machine is not running for a while, here’s the announcement post:


I am now able to deploy to yyz, it seems; however, I’m still getting PCI: Fatal: No config space access function found, [ERROR] [Entrypoint]: mysqld failed while attempting to check config, and mysqld got signal 11 ; Most likely, you have hit a bug, but this error can also be caused by malfunctioning hardware. errors when I try to deploy my MySQL app.

Are any of these likely the result of something on Fly’s end, or just my own stupidity? Thanks!

Ah. thanks @lubien! I wonder if it might be something like that.

My app is back! Thanks so much for your help @andie. I think that it was a bug with the current version of MySQL (8.0.34).

What fixed it was pinning the image to mysql:8.0.32 — which is probably what I was running before the migration — like you (duh) currently recommend in your docs where it warns about version 8.0.33.

(At the same time, I noticed that the fly.toml format for MySQL has changed from [experimental] CMD = … to [processes] app =. I changed that too, but I don’t think that was actually the problem.)


Glad to hear it’s back! And nice catch on figuring out the issue. We have an open ticket for that bug (since 8.0.33).


Is this issue back in CDG?

Getting could not reserve resource for machine: insufficient memory available to fulfill request

1 Like

Getting the same on Paris region.

@Ralph @xantrac CDG and FRA are not at capacity. We should be able to schedule machines there.

Do you use volumes with your apps? Currently volumes are strongly tied to underlying physical hosts and would let our scheduler place your apps in the hosts that have the volumes regardless of the capacity there.

Getting the same on Tokyo region too.

Same in the MAD region

btw why this problem, which is rather critical, is not visible here?

@kitade @mif Hmm, Tokyo (NRT) and MAD are not at capacity. Let me double-check our provisioning logic.

Do you have issues provisioning remote builders, or other apps? Are the apps that have issues have volumes?

Only apps as I can see. Yes, the app does have a volume attached

The most concerning thing is that my app is offline now. I would be ok with not being able to deploy a new version but not by shutting down working version first just to find out that it’s not possible to run the new one :frowning:

Sorry for the inconvenience. There were few out-of-memory emails we sent regarding one of your apps. Could you check your inbox? That might be the reason your app is being offline.

Getting the same on Tokyo region too.
The machine is not responding.
Please take action.
Image from Gyazo

Why there is no incidents in this page??