-
Notifications
You must be signed in to change notification settings - Fork 4.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Can not build for front-proxy of examples #20915
Comments
It works for me, from the log it says it wants to pull |
this seems to be the same issue reported in #20897 im struggling to understand the circumstances under which this is failing - wfm locally, and ci is testing this without issue i think
im inclined to think the same, but the error messages are confusing
im trying to understand how that makes a difference, but we need 3.10 anyway |
Are you behind a corporate proxy? @kakaru1331 You may want to configure your docker to use proxy. |
@bryanwux Yes I am. I think you're right. |
I am running into the same issue I believe,
but without making any changes to docker config locally , I can easily pull these other images
as well as
|
This issue still exists. May I ask is there a plan to fix it? |
i think there is some confusion i see that wont work for im not sure if there is an actual problem, confusion aside, but as said, it works in CI and for me locally, so not sure, but would need a bit more info as to when it occurs if there is |
Oh , sorry , i meant to bring up that docker pull itself was failing but i see what you mean. So, trying verbatim instructions from the front-proxy tutorial and no corporate vpn or firewall or proxy in place
It didnt require any kind of authorization before
trying again with
Tried running a different sandbox (mysql) and everything works properly, no errors.
so, may be the issue is not my local terminal 's connection to dockerhub. docker version
|
ok, thanks, ill go through the sandbox and see if i can repro |
I ran in to the same problem. I do not think it is related python version. It is failing while building tracer image and it is trying to pull I removed |
i suspect it is something like this - the |
same issue May be, problem raised from cannot find envoyproxy:tracing image in docker hub? ref: https://hub.docker.com/search?q=%20envoyproxy%3Atracing |
yep, same issue - altho dockerhub is not really relevant as the images are expected to be built rather than pulled ill try and have a look at this today and see if i can figure why it works in some situations and not others |
thx, docker-compose file ref: https://www.envoyproxy.io/docs/envoy/latest/start/sandboxes/skywalking_tracing |
Encountered same error, I found that the |
dockerhub is not really relevant as the images are expected to be built rather than pulled |
Sorry, I recheck the problem and found it may due to a feature named "buildkit" which is enabled by default for some version of docker engine, you can find more detail here, and you can use |
interesting - apologies i have struggled to find time to follow up on this, ill try to prioritize and check if buildkit is what is causing issue |
I was facing this error when running
In my case, |
Since you specify |
the point is this - it should only pull for images that can be pulled, and build for the image that should be built clearly this is an issue, so will prioritize as it "works" in CI i havent made this a prio so far (apologies) - maybe following the docs never works and only the way its done in CI does - or perhaps something else - will check further... |
so when i follow the steps in the docs for the $ docker-compose pull
Pulling flask_service ... done
Pulling front-envoy ... done
Pulling service1 ... done
Pulling service2 ... done
WARNING: Some service image(s) must be built from source by running:
docker-compose build flask_service
this is expected - as the warning states one of the images - if others are seeing this turn into an error im guessing it must be related to the docker or more liikely docker-compose version please post versions if you are seeing error when following the docs exactly and we can try and figure out when this is happening it should be noted that the |
and then
|
you need to use |
works for me |
ive managed to repro this one of the examples ( |
i think i have a fix for this - i need to test further to confirm but looks promising - i will PR some cleanups for the examples and will include there |
this will be fixed in #21962 for avoidance of doubt, and contrary to what i said previously
apologies if it seemed i doubted - not sure why i didnt manage to repro - either way this should be fixed shortly |
Follow on fix for envoyproxy#20915 Signed-off-by: Ryan Northey <[email protected]>
docs: Fix example build instructions Follow on fix for #20915 Signed-off-by: Ryan Northey <[email protected]>
This is still an issue. Thanks for the easy commands, @dhjack I think the issue can be fixed by specifying the Dockerfile path in docker-compose. Similar to this: https://github.com/grpc/grpc-web/blob/master/docker-compose.yml |
To reproduce, open up: https://gitpod.io/#https://github.com/envoyproxy/envoy and run |
@SunnyAureliusRichard i have just followed those steps (again) and it works as expected - the issue raised here creates the following warning: WARNING: Image for service service2 was built because it did not already exist. To rebuild this image you must use `docker-compose build` or `docker-compose up --build` previously, in some cases this was throwing an error i believe my first question is whether you are running from current Envoy |
the Dockerfile path should not need to be set as the |
hi @phlax . Thanks for taking an interest in this. I've tried a few things, and couldn't get it to work. Turns out that the image "grpcweb/prereqs" mentioned in the link above is in the docker hub as well, therefore their builds don't fail. Their use case seemed very similar to this one.
logs for
|
what version of Envoy are you testing this with (eg commit hash) |
looking through the logs it seems as though this already failed with: panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0xf0ce38] which is not the original error that was raised here im not familiar with gitpod - but im wondering if there is some issue there i would like to reproduce and track down any issue - but i need to know the versions of:
|
sometimes i get that invalid memory address error, it's only happened with this repo. it could be some issue with their system. Unfortunately it's the only dev environment I have access to at the moment, so I can't reproduce it anywhere else. envoy main: 9e7fba8 If you'd like to reproduce it, I've put a link above. |
i have never seen this issue before so i think it is more to do with gitpod
i have tried reproducing and it works as expected in my env could you try and reproduce the problem somewhere other than on gitpod
this is a |
Getting the same error. MacOS 12.6 on M1 using Docker Desktop. Head:
Tried to logout and login to docker hub, same results. |
envoy/examples/shared/flask/Dockerfile specifies a specific version of the python container with |
Hi, I hit this issue also. I followed the instructions here:(https://www.envoyproxy.io/docs/envoy/latest/start/sandboxes/front_proxy.html).
Running Thanks! |
it appears your this is known to not work atm its on my radar to switch the examples -> |
heads up to anyone tracking the issues discussed here ... we have finally switched the examples/documentation to using the i believe that the examples should still work with the python the documentation is still being updated to reflect the changes, feedback, prs etc appreciated |
Can't build front-proxy example with docker-compose.
It has been failed when pulling python images used in shared resources (flask and tracing).
And I have solved to change the image from 3.10-slim-bullseye to 3.11.0a7-bullseye
I think it could be a docker registry problem rather than the example.
I'll register the issue just in case.
And the detailed error log is below.
=> ERROR [internal] load metadata for docker.io/library/flask_service:python-3.10-slim-bullseye
failed to solve with frontend dockerfile.v0: failed to create LLB definition: pull access denied, repository does not exist or may require authorization: server message: insufficient_scope: authorizatio
ERROR: Service 'tracing' failed to build : Build failed
The text was updated successfully, but these errors were encountered: