I can see there's a grayed out option to assign groups to projects as members, but I can't find a place to create the groups in the UI.

Assigning individual users to projects is pretty painful.

I'm using the local database as the authentication source.

Harbor Version v1.10.2-d0189bed

Might need some changes like
uber/makisu#223

The deploy_test and queue_test go files contain a lot of repetitive and duplicated code that should be extracted into helpers.
My guesstimate is that we can save probably some 30-40% of the files. And they are more than 2k lines together. This is a sizable readability improvement.

In what area(s)?

/area API
/area test-and-release

/kind good-first-issue

What would you like to be added/modified:
For a device mapper, currently the configmap would be automatically generated by device controller, but users need to manually indicated the configmap name in podSpec, which doesn't make sence.

We can simply make admission controller to auto inject devicemapper relevant c

If it already is, it serves two masters (which isn't clear).

e.g. better documentation for self-signed certs

Kata shimv2 does not have any tracing yet. Will be useful to add this.

Either @mattmoor creates github.com/mattmoor/examples or we change the import paths to something that actually exists (e.g. ./cmd/ko/test).

What are you trying to do?
I would like to be able to use a cloud-init.yml from a url while launching a new vm

multipass launch -n test -c 2 --cloud-init http://my-home.com/multipass/cloud-init.yml

What's your proposed solution?
--cloud-init should support http:// files

Additional context
Add any other context or screenshots about the feature request here.

Version of Singularity:

$ singularity version
3.5.2

Expected behavior

Given this simple def file:

$ more simple.def 
Bootstrap: library
From: ubuntu:18.04

%post

    mkdir -p /opt/openmpi-4.0.1 && cd $_
    pwd 

The pwd should print /opt/openmpi-4.0.1

Actual behavior

S

Is your feature request related to a problem?
"DevSpace should use the current docker behavior for dockerignore:
Dockerfile + .dockerignore
Dockerfile + Dockerfile.dockerignore
Dockerfile.frontend + Dockerfile.frontend.dockerignore
"

Which solution do you suggest?
Delegating the dockerignore file search to Docker if possible, otherwise copying the behavior as best as possible

Describe the bug
I just signed up and went to the public templates section.
There is only one template shown (see screenshot).

To Reproduce
Steps to reproduce the behavior:

1. Go to 'https://cloud.lastbackend.com/templates#public'
2. See error

Expected behavior
I was expecting a lot more, e.g. Wordpress, MongoDB, Bootstrap, PostgreSQL, Etherpad, Apache, CouchDB, SearX, wall

Many advanced K8s workload management features require the use of labels. In some cases, labels are needed after Pods are created so that adding them to Pod template does not help . For example,

• A canary release plan may want to upgrade X number of Pods out of Y number of total replica. We can add a label to the chosen Pod so that the rollout controller can upgrade these Pods first.

• Po

Hi, great work - saved me from a lot of pain!

Quick question - as I was using nginx amplify before, how can I integrate it here?

e.g. using the Dockerfile from https://github.com/nginxinc/docker-nginx-amplify

FROM nginx:latest

RUN apt-get update \
    && apt-get install -qqy curl python apt-transport-https apt-utils gnupg1 procps \
    && echo 'deb https://packages.amplify.nginx

Describe the bug
函数多instance及timeout特性不符合预期

To Reproduce
某python函数配置多个实例以及timeout，配置如下

向source topic中依次发送4条消息
消息1：异常消息，发送后能够触发函数除0错误
消息2：异常消息，发送后触发函数执行超时
消息3：异常消息，发送后触发函数执行超时
消息4：正常消息，发送后能有正常返回

Expected behavior
函数最大实例为3，预期发完4条消息后能立马收到消息1和消息4的执行结果
然后超时后，也就是10s

At the top of the docs it mentions lazy-loading, but I can't find anything in the docs regarding where and how is this available.

Sample use-cases would be nice as well :-)

HI, I'm trying to perform the migration between to VMs (with Xubuntu inside) hosted by Virtualbox. At the beginning I was doing it with LXD "move" but only cold migration worked. So I tried to make a snapshot and this is what I got.

lxc snapshot prova1 sn --stateful
lxc stop prova1 --stateful
lxc start prova1
Error: Migrate: Failed to run: /snap/lxd/current/bin/lxd forkmigrate prova1 /

title might be wrong, here's what I'm doing

@singleton()
export default class StripeConnectionTokenResolversFactory extends AbstractResolversFactory {

@registry([
  { token: InjectToken.GraphFieldResolvers, useToken: StripeConnectionTokenResolversFactory },
  { token: InjectToken.ApolloProvider, useClass: ApolloProvider },
  { token: InjectToken.ContentfulProvider, use

Architecture as well as workflows tend to be very complex within the appserver.io environment.
We need diagrams for better documentation and overview of architecture and request/response flow.

UAC:

• There MUST be workflow diagrams showing the request/response flow
• Diagrams MUST illustrate the difference in between servlet requests and legacy ones
• There MUST be architectural diagrams
• At lea

我修改了config.filebeat
cat >> $FILEBEAT_CONFIG << EOF
$(base)
output.elasticsearch:
hosts: [$ELASTICSEARCH_HOSTS]
index: ${ELASTICSEARCH_INDEX:-filebeat}-%{+yyyy.MM}
${ELASTICSEARCH_SCHEME:+protocol: ${ELASTICSEARCH_SCHEME}}
${ELASTICSEARCH_USER:+username: ${ELASTICSEARCH_USER}}
${ELASTICSEARCH_PASSWORD:+password: ${ELASTICSEARCH_PASSWORD}}
${ELASTICSEARCH_WORKER:+wor

I was just trying to build caddy binary with a custom set of plugins, but got stuck with the invalid import path "" error.

Looks like most of us are trying to use the qualified name of the plugin instead of just the last part of it (abiosoft/caddy-docker#108 (comment)). I.e:

RUN VERSION="1.0.3" \
    PLUGINS="http.prometheus" \
    ENABLE_TELE

Docker has this for pull: https://docs.docker.com/engine/reference/commandline/pull/

The tarball format supports multiple images/tags. We'd just want to pull every tag for crane but with gcrane we could pull all the untagged manifests as well.

Docker doesn't have it for push but I don't see a reason not to support it.

This would work well with an --oci flag to do an image layout ins