Upstream/harbor
1
0
Fork 0
mirror of https://github.com/goharbor/harbor.git synced 2024-06-29 00:05:09 +02:00
Code Issues Projects Releases Wiki Activity

Merge branch 'master' into job-service

Browse Source
This commit is contained in:
Tan Jiang 2016-05-19 16:12:21 +08:00
commit fcd0de0ca0
17 changed files with 359 additions and 176 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
AUTHORS
View File

@ -1,17 +1,27 @@
# This file lists all individuals having contributed content to the repository.
Alexander Zeitler <alexander.zeitler at pdmlab.com>
Alexey Erkak <eryigin at mail.ru>
Allen Heavey <xheavey at gmail.com>
Amanda Zhang <amzhang at vmware.com>
Benniu Ji <benniuji at gmail.com>
Bobby Zhang <junzhang at vmware.com>
Chaofeng Wu <chaofengw at vmware.com>
Daniel Jiang <jiangd at vmware.com>
Deshi Xiao <xiaods at gmail.com>
Guangping Fu <krystism at gmail.com>
Haining Henry Zhang <henryzhang at vmware.com>
Hao Xia <haox at vmware.com>
Jack Liu <ljack at vmware.com>
Jessy Zhang <jessyz at vmware.com>
Kun Wang <kunw at vmware.com>
Mahesh Paolini-Subramanya <mahesh at dieswaytoofast.com>
Meng Wei <weim at vmware.com>
Nagarjun G <nagarjung.g at gmail.com>
Peng Zhao <zhaopeng1988 at gmail.com>
Robin Naundorf <r.naundorf at fh-muenster.de>
Shan Zhu <zhus at vmware.com>
Victoria Zheng <vzheng at vmware.com>
Wenkai Yin <yinw at vmware.com>
Yahao He <bhe at vmware.com>
Yan Wang <wangyan at vmware.com>

14
README.md
View File

@ -9,18 +9,18 @@
Project Harbor is an enterprise-class registry server, which extends the open source Docker Registry server by adding the functionality usually required by an enterprise, such as security, control, and management. Harbor is primarily designed to be a private registry - providing the needed security and control that enterprises require. It also helps minimize bandwidth usage, which is helpful to both improve productivity (local network access) as well as performance (for those with poor internet connectivity).
### Features
* **Role Based Access Control**: Users and docker repositories are organized via "projects", a user can have different permission for images under a namespace.
* **Graphical user portal**: User can easily browse, search docker repositories, manage projects/namespaces.
* **Role Based Access Control**: Users and Docker repositories are organized via "projects", a user can have different permission for images under a project.
* **Graphical user portal**: User can easily browse, search Docker repositories, manage projects/namespaces.
* **AD/LDAP support**: Harbor integrates with existing enterprise AD/LDAP for user authentication and management.
* **Auditing**: All the operations to the repositories are tracked.
* **Internationalization**: Already Localized for English, Chinese and German. More languages can be added.
* **Internationalization**: Already localized for English, Chinese, German and Russian. More languages can be added.
* **RESTful API**: RESTful APIs for most administrative operations, easing intergration with external management platforms.
### Getting Started
Harbor is self-contained and can be easily deployed via docker-compose (Quick-Start steps below). Refer to the [Installation and Configuration Guide](docs/installation_guide.md) for detailed information.
**System requirements:**
Harbor only works with docker 1.10+ and docker-compose 1.6.0+, and an internet-connected host
Harbor only works with docker 1.10+ and docker-compose 1.6.0+, and an internet-connected host.
1. Get the source code:
@ -30,7 +30,7 @@ Harbor only works with docker 1.10+ and docker-compose 1.6.0+, and an internet-c
2. Edit the file **Deploy/harbor.cfg**, make necessary configuration changes such as hostname, admin password and mail server. Refer to [Installation and Configuration Guide](docs/installation_guide.md) for more info.
3. Install Harbor with the following commands. Note that the docker-compose process can take a while!
3. Install Harbor with the following commands. Note that the docker-compose process can take a while.
```sh
$ cd Deploy
@ -56,9 +56,6 @@ For those who don't want to clone the source, or need to install Harbor on a ser
For information on how to use Harbor, please see [User Guide](docs/user_guide.md) .
### Deploy Harbor on Kubernetes
Detailed instruction about deploying Harbor on Kubernetes is available [here](docs/kubernetes_deployment.md).
### Contribution
We welcome contributions from the community. If you wish to contribute code and you have not signed our contributor license agreement (CLA), our bot will update the issue when you open a pull request. For any questions about the CLA process, please refer to our [FAQ](https://cla.vmware.com/faq).
@ -67,6 +64,7 @@ Harbor is available under the [Apache 2 license](LICENSE).
### Partners
<a href="https://www.shurenyun.com/" border="0" target="_blank"><img alt="DataMan" src="docs/img/dataman.png"></a> &nbsp; &nbsp; <a href="http://www.slamtec.com" target="_blank" border="0"><img alt="SlamTec" src="docs/img/slamteclogo.png"></a>
&nbsp; &nbsp; <a href="https://www.caicloud.io" border="0"><img alt="CaiCloud" src="docs/img/caicloudLogoWeb.png"></a>
### Users
<a href="https://www.madailicai.com/" border="0" target="_blank"><img alt="MaDaiLiCai" src="docs/img/UserMaDai.jpg"></a>

45
api/project.go
View File

@ -113,23 +113,50 @@ func (p *ProjectAPI) Head() {
// Get ...
func (p *ProjectAPI) Get() {
queryProject := models.Project{UserID: p.userID}
var projectList []models.Project
projectName := p.GetString("project_name")
if len(projectName) > 0 {
queryProject.Name = "%" + projectName + "%"
projectName = "%" + projectName + "%"
}
var public int
var err error
isPublic := p.GetString("is_public")
if len(isPublic) > 0 {
public, err = strconv.Atoi(isPublic)
if err != nil {
log.Errorf("Error parsing public property: %d, error: %v", isPublic, err)
p.CustomAbort(http.StatusBadRequest, "invalid project Id")
}
}
isAdmin := false
if public == 1 {
projectList, err = dao.GetPublicProjects(projectName)
} else {
isAdmin, err = dao.IsAdminRole(p.userID)
if err != nil {
log.Errorf("Error occured in check admin, error: %v", err)
p.CustomAbort(http.StatusInternalServerError, "Internal error.")
}
if isAdmin {
projectList, err = dao.GetAllProjects(projectName)
} else {
projectList, err = dao.GetUserRelevantProjects(p.userID, projectName)
}
}
public, _ := p.GetInt("is_public")
queryProject.Public = public
projectList, err := dao.QueryProject(queryProject)
if err != nil {
log.Errorf("Error occurred in QueryProject, error: %v", err)
log.Errorf("Error occured in get projects info, error: %v", err)
p.CustomAbort(http.StatusInternalServerError, "Internal error.")
}
for i := 0; i < len(projectList); i++ {
if isProjectAdmin(p.userID, projectList[i].ProjectID) {
projectList[i].Togglable = true
if public != 1 {
if isAdmin {
projectList[i].Role = models.PROJECTADMIN
}
if projectList[i].Role == models.PROJECTADMIN {
projectList[i].Togglable = true
}
}
projectList[i].RepoCount = getRepoCountByProject(projectList[i].Name)
}
p.Data["json"] = projectList
p.ServeJSON()

4
api/search.go
View File

@ -55,13 +55,13 @@ func (s *SearchAPI) Get() {
var projects []models.Project
if isSysAdmin {
projects, err = dao.GetAllProjects()
projects, err = dao.GetAllProjects("")
if err != nil {
log.Errorf("failed to get all projects: %v", err)
s.CustomAbort(http.StatusInternalServerError, "internal error")
}
} else {
projects, err = dao.GetUserRelevantProjects(userID)
projects, err = dao.SearchProjects(userID)
if err != nil {
log.Errorf("failed to get user %d 's relevant projects: %v", userID, err)
s.CustomAbort(http.StatusInternalServerError, "internal error")

117
api/statistic.go Normal file
View File

@ -0,0 +1,117 @@
/*
Copyright (c) 2016 VMware, Inc. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
*/
package api
import (
"net/http"
"strings"
"github.com/vmware/harbor/dao"
"github.com/vmware/harbor/models"
svc_utils "github.com/vmware/harbor/service/utils"
"github.com/vmware/harbor/utils/log"
)
// StatisticAPI handles request to /api/statistics/
type StatisticAPI struct {
BaseAPI
userID int
}
//Prepare validates the URL and the user
func (s *StatisticAPI) Prepare() {
s.userID = s.ValidateUser()
}
// Get total projects and repos of the user
func (s *StatisticAPI) Get() {
isAdmin, err := dao.IsAdminRole(s.userID)
if err != nil {
log.Errorf("Error occured in check admin, error: %v", err)
s.CustomAbort(http.StatusInternalServerError, "Internal error.")
}
var projectList []models.Project
if isAdmin {
projectList, err = dao.GetAllProjects("")
} else {
projectList, err = dao.GetUserRelevantProjects(s.userID, "")
}
if err != nil {
log.Errorf("Error occured in QueryProject, error: %v", err)
s.CustomAbort(http.StatusInternalServerError, "Internal error.")
}
proMap := map[string]int{}
proMap["my_project_count"] = 0
proMap["my_repo_count"] = 0
proMap["public_project_count"] = 0
proMap["public_repo_count"] = 0
var publicProjects []models.Project
publicProjects, err = dao.GetPublicProjects("")
if err != nil {
log.Errorf("Error occured in QueryPublicProject, error: %v", err)
s.CustomAbort(http.StatusInternalServerError, "Internal error.")
}
proMap["public_project_count"] = len(publicProjects)
for i := 0; i < len(publicProjects); i++ {
proMap["public_repo_count"] += getRepoCountByProject(publicProjects[i].Name)
}
if isAdmin {
proMap["total_project_count"] = len(projectList)
proMap["total_repo_count"] = getTotalRepoCount()
}
for i := 0; i < len(projectList); i++ {
if isAdmin {
projectList[i].Role = models.PROJECTADMIN
}
if projectList[i].Role == models.PROJECTADMIN || projectList[i].Role == models.DEVELOPER ||
projectList[i].Role == models.GUEST {
proMap["my_project_count"]++
proMap["my_repo_count"] += getRepoCountByProject(projectList[i].Name)
}
}
s.Data["json"] = proMap
s.ServeJSON()
}
//getReposByProject returns repo numbers of specified project
func getRepoCountByProject(projectName string) int {
repoList, err := svc_utils.GetRepoFromCache()
if err != nil {
log.Errorf("Failed to get repo from cache, error: %v", err)
return 0
}
var resp int
if len(projectName) > 0 {
for _, r := range repoList {
if strings.Contains(r, "/") && r[0:strings.LastIndex(r, "/")] == projectName {
resp++
}
}
return resp
}
return 0
}
//getTotalRepoCount returns total repo count
func getTotalRepoCount() int {
repoList, err := svc_utils.GetRepoFromCache()
if err != nil {
log.Errorf("Failed to get repo from cache, error: %v", err)
return 0
}
return len(repoList)
}

30
contrib/Configure_mirror.md Normal file
View File

@ -0,0 +1,30 @@
# Configuring Harbor as a local registry mirror
Harbor runs as a local registry by default. It can also be configured as a registry mirror,
which caches downloaded images for subsequent use. Note that under this setup, the Harbor registry only acts as a mirror server and
no longer accepts image pushing requests. Edit `Deploy/templates/registry/config.yml` before executing `./prepare`, and append a `proxy` section as follows:
```
proxy:
remoteurl: https://registry-1.docker.io
```
In order to access private images on the Docker Hub, a username and a password can be supplied:
```
proxy:
remoteurl: https://registry-1.docker.io
username: [username]
password: [password]
```
You will need to pass the `--registry-mirror` option to your Docker daemon on startup:
```
docker --registry-mirror=https://<my-docker-mirror-host> daemon
```
For example, if your mirror is serving on `http://reg.yourdomain.com`, you would run:
```
docker --registry-mirror=https://reg.yourdomain.com daemon
```
Refer to the [Registry as a pull through cache](https://github.com/docker/distribution/blob/master/docs/mirror.md) for detailed information.

2
contrib/README.md
View File

@ -1 +1 @@
The `contrib` directory contains documents, scripts, and other helpful things which are contributed by community.
The `contrib` directory contains documents, scripts, and other helpful things which are contributed by the community.

58
dao/dao_test.go
View File

@ -351,7 +351,7 @@ func TestChangeUserPasswordWithIncorrectOldPassword(t *testing.T) {
}
func TestQueryRelevantProjectsWhenNoProjectAdded(t *testing.T) {
projects, err := GetUserRelevantProjects(currentUser.UserID)
projects, err := SearchProjects(currentUser.UserID)
if err != nil {
t.Errorf("Error occurred in QueryRelevantProjects: %v", err)
}
@ -570,39 +570,6 @@ func TestIsProjectPublic(t *testing.T) {
}
}
func TestQueryProject(t *testing.T) {
query1 := models.Project{
UserID: 1,
}
projects, err := QueryProject(query1)
if err != nil {
t.Errorf("Error in Query Project: %v, query: %+v", err, query1)
}
if len(projects) != 2 {
t.Errorf("Expecting get 2 projects, but actual: %d, the list: %+v", len(projects), projects)
}
query2 := models.Project{
Public: 1,
}
projects, err = QueryProject(query2)
if err != nil {
t.Errorf("Error in Query Project: %v, query: %+v", err, query2)
}
if len(projects) != 1 {
t.Errorf("Expecting get 1 project, but actual: %d, the list: %+v", len(projects), projects)
}
query3 := models.Project{
UserID: 9,
}
projects, err = QueryProject(query3)
if err != nil {
t.Errorf("Error in Query Project: %v, query: %+v", err, query3)
}
if len(projects) != 0 {
t.Errorf("Expecting get 0 project, but actual: %d, the list: %+v", len(projects), projects)
}
}
func TestGetUserProjectRoles(t *testing.T) {
r, err := GetUserProjectRoles(currentUser.UserID, currentProject.ProjectID)
if err != nil {
@ -630,20 +597,20 @@ func TestProjectPermission(t *testing.T) {
}
func TestGetUserRelevantProjects(t *testing.T) {
projects, err := GetUserRelevantProjects(currentUser.UserID)
projects, err := GetUserRelevantProjects(currentUser.UserID, "")
if err != nil {
t.Errorf("Error occurred in GetUserRelevantProjects: %v", err)
}
if len(projects) != 2 {
t.Errorf("Expected length of relevant projects is 2, but actual: %d, the projects: %+v", len(projects), projects)
if len(projects) != 1 {
t.Errorf("Expected length of relevant projects is 1, but actual: %d, the projects: %+v", len(projects), projects)
}
if projects[1].Name != projectName {
if projects[0].Name != projectName {
t.Errorf("Expected project name in the list: %s, actual: %s", projectName, projects[1].Name)
}
}
func TestGetAllProjects(t *testing.T) {
projects, err := GetAllProjects()
projects, err := GetAllProjects("")
if err != nil {
t.Errorf("Error occurred in GetAllProjects: %v", err)
}
@ -655,6 +622,19 @@ func TestGetAllProjects(t *testing.T) {
}
}
func TestGetPublicProjects(t *testing.T) {
projects, err := GetPublicProjects("")
if err != nil {
t.Errorf("Error occurred in getProjects: %v", err)
}
if len(projects) != 1 {
t.Errorf("Expected length of projects is 1, but actual: %d, the projects: %+v", len(projects), projects)
}
if projects[0].Name != "library" {
t.Errorf("Expected project name in the list: %s, actual: %s", "library", projects[0].Name)
}
}
func TestAddProjectMember(t *testing.T) {
err := AddProjectMember(currentProject.ProjectID, 1, models.DEVELOPER)
if err != nil {

100
dao/project.go
View File

@ -79,42 +79,6 @@ func IsProjectPublic(projectName string) bool {
return project.Public == 1
}
// QueryProject querys the projects based on publicity and user, disregarding the names etc.
func QueryProject(query models.Project) ([]models.Project, error) {
o := orm.NewOrm()
sql := `select distinct
p.project_id, p.owner_id, p.name,p.creation_time, p.update_time, p.public
from project p
left join project_member pm on p.project_id = pm.project_id
where p.deleted = 0 `
queryParam := make([]interface{}, 1)
if query.Public == 1 {
sql += ` and p.public = ?`
queryParam = append(queryParam, query.Public)
} else if isAdmin, _ := IsAdminRole(query.UserID); isAdmin == false {
sql += ` and (pm.user_id = ?) `
queryParam = append(queryParam, query.UserID)
}
if query.Name != "" {
sql += " and p.name like ? "
queryParam = append(queryParam, query.Name)
}
sql += " order by p.name "
var r []models.Project
_, err := o.Raw(sql, queryParam).QueryRows(&r)
if err != nil {
return nil, err
}
return r, nil
}
//ProjectExists returns whether the project exists according to its name of ID.
func ProjectExists(nameOrID interface{}) (bool, error) {
o := orm.NewOrm()
@ -208,11 +172,11 @@ func ToggleProjectPublicity(projectID int64, publicity int) error {
return err
}
// GetUserRelevantProjects returns a project list,
// SearchProjects returns a project list,
// which satisfies the following conditions:
// 1. the project is not deleted
// 2. the prject is public or the user is a member of the project
func GetUserRelevantProjects(userID int) ([]models.Project, error) {
func SearchProjects(userID int) ([]models.Project, error) {
o := orm.NewOrm()
sql := `select distinct p.project_id, p.name, p.public
from project p
@ -228,14 +192,66 @@ func GetUserRelevantProjects(userID int) ([]models.Project, error) {
return projects, nil
}
// GetAllProjects returns all projects which are not deleted
func GetAllProjects() ([]models.Project, error) {
// GetUserRelevantProjects returns the projects of the user which are not deleted and name like projectName
func GetUserRelevantProjects(userID int, projectName string) ([]models.Project, error) {
o := orm.NewOrm()
sql := `select project_id, name, public
sql := `select distinct
p.project_id, p.owner_id, p.name,p.creation_time, p.update_time, p.public, pm.role role
from project p
left join project_member pm on p.project_id = pm.project_id
where p.deleted = 0 and pm.user_id= ?`
queryParam := make([]interface{}, 1)
queryParam = append(queryParam, userID)
if projectName != "" {
sql += " and p.name like ? "
queryParam = append(queryParam, projectName)
}
sql += " order by p.name "
var r []models.Project
_, err := o.Raw(sql, queryParam).QueryRows(&r)
if err != nil {
return nil, err
}
return r, nil
}
//GetPublicProjects returns all public projects whose name like projectName
func GetPublicProjects(projectName string) ([]models.Project, error) {
publicProjects, err := getProjects(1, projectName)
if err != nil {
return nil, err
}
return publicProjects, nil
}
// GetAllProjects returns all projects which are not deleted and name like projectName
func GetAllProjects(projectName string) ([]models.Project, error) {
allProjects, err := getProjects(0, projectName)
if err != nil {
return nil, err
}
return allProjects, nil
}
func getProjects(public int, projectName string) ([]models.Project, error) {
o := orm.NewOrm()
sql := `select project_id, owner_id, creation_time, update_time, name, public
from project
where deleted = 0`
queryParam := make([]interface{}, 1)
if public == 1 {
sql += " and public = ? "
queryParam = append(queryParam, public)
}
if len(projectName) > 0 {
sql += " and name like ? "
queryParam = append(queryParam, projectName)
}
sql += " order by name "
var projects []models.Project
if _, err := o.Raw(sql).QueryRows(&projects); err != nil {
log.Debugf("sql xxx", sql)
if _, err := o.Raw(sql, queryParam).QueryRows(&projects); err != nil {
return nil, err
}
return projects, nil

BIN
docs/img/caicloudLogoWeb.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 6.9 KiB

113
docs/installation_guide.md
View File

@ -4,7 +4,9 @@ Harbor can be installed in one of two ways:
1. From source code - This goes through a full build process, _and requires an Internet connection_.
2. Pre-built installation package - This can save time (no building necessary!) as well as allows for installation on a host that is _not_ connected to the Internet.
This guide describes both of these approaches
This guide describes both of these approaches.
In addition, the deployment instructions on Kubernetes has been created by the community. Refer to [Deploy Harbor on Kubernetes](kubernetes_deployment.md) for details.
## Prerequisites for the target host
Harbor is deployed as several Docker containers, and, therefore, can be deployed on any Linux distribution that supports Docker.
@ -35,7 +37,7 @@ The parameters are described below - note that at the very least, you will need
* **hostname**: The target host's hostname, which is used to access the UI and the registry service. It should be the IP address or the fully qualified domain name (FQDN) of your target machine, e.g., `192.168.1.10` or `reg.yourdomain.com`. _Do NOT use `localhost` or `127.0.0.1` for the hostname - the registry service needs to be accessible by external clients!_
* **ui_url_protocol**: (**http** or **https**. Default is **http**) The protocol used to access the UI and the token/notification service. By default, this is _http_. To set up the https protocol, refer to [Configuring Harbor with HTTPS Access](configure_https.md).
* **Email settings**: These parameters are needed for Harbor to be able to send a user a "password reset" email, and are only necessary if that functionality is needed. Also, do mnote that by default SSL connectivity is _not_ enabled - if your SMTP server requires SSL, but does _not_ support STARTTLS, then you should enable SSL by setting **email_ssl = true**.
* **Email settings**: These parameters are needed for Harbor to be able to send a user a "password reset" email, and are only necessary if that functionality is needed. Also, do note that by default SSL connectivity is _not_ enabled - if your SMTP server requires SSL, but does _not_ support STARTTLS, then you should enable SSL by setting **email_ssl = true**.
* email_server = smtp.mydomain.com
* email_server_port = 25
* email_username = sample_admin@mydomain.com
@ -47,11 +49,33 @@ The parameters are described below - note that at the very least, you will need
* **auth_mode**: The type of authentication that is used. By default it is **db_auth**, i.e. the credentials are stored in a database. For LDAP authentication, set this to **ldap_auth**.
* **ldap_url**: The LDAP endpoint URL (e.g. `ldaps://ldap.mydomain.com`). _Only used when **auth_mode** is set to *ldap_auth* ._
* **ldap_basedn**: The basedn template for verifying the user's credentials against LDAP (e.g. `uid=%s,ou=people,dc=mydomain,dc=com`). _Only used when **auth_mode** is set to *ldap_auth* ._
* **db_password**: The root password for the mySQL database used for **db_auth**. _Change this password for any production use!!_
* **self_registration**: (**on** or **off**. Default is **on**) Enable / Disable the ability for a user to register themselves. When disabled, new users can only be created by the Admin user, only an admin user can create new users in Harbor. _NOTE: When **auth_mode** is set to **ldap_auth**, self-registration feature is **always** disabled, and this flag is ignored.
* **db_password**: The root password for the mySQL database used for **db_auth**. _Change this password for any production use!_
* **self_registration**: (**on** or **off**. Default is **on**) Enable / Disable the ability for a user to register themselves. When disabled, new users can only be created by the Admin user, only an admin user can create new users in Harbor. _NOTE: When **auth_mode** is set to **ldap_auth**, self-registration feature is **always** disabled, and this flag is ignored._
#### Configuring storage backend (optional)
By default, Harbor stores images on your local filesystem. In a production environment, you may consider
using other storage backend instead of the local filesystem, like S3, Openstack Swift, Ceph, etc.
What you need to update is the section of `storage` in the file `Deploy/templates/registry/config.yml`.
For example, if you use Openstack Swift as your storage backend, the section may look like this:
```
storage:
swift:
username: admin
password: ADMIN_PASS
authurl: http://keystone_addr:35357/v3
tenant: admin
domain: default
region: regionOne
container: docker_images
```
_NOTE: For detailed information on storage backend of a registry, refer to [Registry Configuration Reference](https://docs.docker.com/registry/configuration/) ._
#### Building and starting Harbor
Once **harbord.cfg** is configured, build and start Harbor as follows. Note that Note that the docker-compose process can take a while!
Once **harbord.cfg** and storage backend (optional) are configured, build and start Harbor as follows. Note that the docker-compose process can take a while.
```sh
$ cd Deploy
@ -77,52 +101,9 @@ $ docker push reg.yourdomain.com/myproject/myrepo
For information on how to use Harbor, please refer to [User Guide of Harbor](user_guide.md) .
#### Configuring Harbor with HTTPS Access
Harbor does not ship with any certificates, and, by default, uses HTTP to serve requests. While this makes it relatively simple to set-up and run - especially for a development or testing environment - it is **not** recommended for a production environment. To enable HTTPS, please refer to [Configuring Harbor with HTTPS Access](configure_https.md)
#### Configuring Harbor with HTTPS access
Harbor does not ship with any certificates, and, by default, uses HTTP to serve requests. While this makes it relatively simple to set up and run - especially for a development or testing environment - it is **not** recommended for a production environment. To enable HTTPS, please refer to [Configuring Harbor with HTTPS Access](configure_https.md).
#### Configuring Harbor as a local registry mirror
The Harbor runs as a local private registry by default, it can be easily configured to run as a local registry mirror, which can keep most of the redundant image fetch traffic on your local network. You just need to edit `config/registry/config.yml` after execute `./prepare`, and append a `proxy` section as follows:
```
proxy:
remoteurl: https://registry-1.docker.io
```
In order to access private images on the Docker Hub, a username and password can be supplied:
```
proxy:
remoteurl: https://registry-1.docker.io
username: [username]
password: [password]
```
You will need to pass the `--registry-mirror` option to your Docker daemon on startup:
```
docker --registry-mirror=https://<my-docker-mirror-host> daemon
```
For example, if your mirror is serving on `http:/reg.yourdomain.com`, you would run:
```
docker --registry-mirror=https://reg.yourdomain.com daemon
```
Refer to the [Registry as a pull through cache](https://github.com/docker/distribution/blob/master/docs/mirror.md) for detail information.
#### Configuring storage backend
By default, the Harbor store images on your local filesystem. In production environment, you may consider using higher available storage backend instead of the local filesystem, like S3, Openstack Swift, Ceph, etc. Fortunately, the Registry supports multiple storage backend, refer to the [Registry Configuration Reference](https://docs.docker.com/registry/configuration/) for detail information. All you need to do is update the section of `storage`, and fill in the fields according to your specied backend. For example, if you use Openstack Swift as your storage backend, the file may look like this:
```
storage:
swift:
username: admin
password: ADMIN_PASS
authurl: http://keystone_addr:35357/v3
tenant: admin
domain: default
region: regionOne
container: docker_images
```
## Installation from a pre-built package
@ -135,7 +116,7 @@ $ cd harbor
Next, configure Harbor as described earlier in [Configuring Harbor](#configuring-harbor).
Finally, run the **prepare** script to generate config files, and use docker compose to build / start Harbor.
Finally, run the **prepare** script to generate config files, and use docker compose to build and start Harbor.
```
@ -151,23 +132,25 @@ $ sudo docker-compose up -d
```
### Deploying Harbor on a host which does not have Internet access
*docker-compose up* pulls the base images from Docker Hub and builds new images for the containers, which, necessarily, requires internet access. To deploy Harbor on a host that is not connected to the Internet
*docker-compose up* pulls the base images from Docker Hub and builds new images for the containers, which, necessarily, requires Internet access. To deploy Harbor on a host that is not connected to the Internet:
1. Prepare Harbor on a machine that has access to the Internet.
2. Export the images as tgz files
3. Transfer them to the target host.
4. Load the tgz file into Docker's local image repo on the host.
THese steps are detailed below
These steps are detailed below:
#### Building and saving images for offline installation
On a machine that is connected to the Internet,
On a machine that is connected to the Internet,
1. Extract the files from the pre-built installation package.
2. Then, run `docker-compose build` to build the images.
3. Use the script `save_image.sh` to export these images as tar files. Note that the tar files will be stored in the `images/` directory.
4. Package everything in the directory `harbor/` into a tgz file
5. Transfer this tgz file to the target machine.
The commands, in detail, are as follows
The commands, in detail, are as follows:
```
$ cd harbor
@ -188,8 +171,8 @@ $ cd ../
$ tar -cvzf harbor_offline-0.1.1.tgz harbor
```
The file `harbor_offline-0.1.1.tgz` contains the images and other files required to start Harbor. You can use tools such as `rsync` or `scp` to transfer the this file to the target host.
On the target host, execute the following commands to start Harbor. _Note that before running the **prepare** script, you **must** update **harbor.cfg** to reflect the right configuration of the target machine!!_ (Refer to Section [Configuring Harbor](#configuring-harbor)
The file `harbor_offline-0.1.1.tgz` contains the images and other files required to start Harbor. You can use tools such as `rsync` or `scp` to transfer this file to the target host.
On the target host, execute the following commands to start Harbor. _Note that before running the **prepare** script, you **must** update **harbor.cfg** to reflect the right configuration of the target machine!_ (Refer to Section [Configuring Harbor](#configuring-harbor)).
```
$ tar -xzvf harbor_offline-0.1.1.tgz
@ -219,7 +202,7 @@ $ sudo docker-compose up -d
```
### Managing Harbor's lifecycle
You can use docker-compose to manage the container lifecycle of the containers. A few useful commands are listed below:
You can use docker-compose to manage the lifecycle of the containers. A few useful commands are listed below:
*Build and start Harbor:*
```
@ -239,7 +222,7 @@ Stopping harbor_registry_1 ... done
Stopping harbor_mysql_1 ... done
Stopping harbor_log_1 ... done
```
*Restart Harbor after stopping*
*Restart Harbor after stopping:*
```
$ sudo docker-compose start
Starting harbor_log_1
@ -247,8 +230,8 @@ Starting harbor_mysql_1
Starting harbor_registry_1
Starting harbor_ui_1
Starting harbor_proxy_1
````
*Remove Harbor's containers while keeping the image data and Harbor's database files on the file system: *
```
*Remove Harbor's containers while keeping the image data and Harbor's database files on the file system:*
```
$ sudo docker-compose rm
Going to remove harbor_proxy_1, harbor_ui_1, harbor_registry_1, harbor_mysql_1, harbor_log_1
@ -265,14 +248,14 @@ $ rm -r /data/database
$ rm -r /data/registry
```
Please check the [Docker Compose command-line reference](https://docs.docker.com/compose/reference/) for more on docker-compose
Please check the [Docker Compose command-line reference](https://docs.docker.com/compose/reference/) for more on docker-compose.
### Persistent data and log files
By default, registry data is persisted in the target host's `/data/` of directory. This data remains unchanged even when Harbor's containers are removed and/or recreated.
In addition, Harbor users `rsyslog` to collect the logs of each container. By default, these log files are stored in the directory `/var/log/harbor/` on the target host.
By default, registry data is persisted in the target host's `/data/` directory. This data remains unchanged even when Harbor's containers are removed and/or recreated.
In addition, Harbor uses `rsyslog` to collect the logs of each container. By default, these log files are stored in the directory `/var/log/harbor/` on the target host.
##Troubleshooting
1.When setting up Harbor behind an nginx proxy or elastic load balancing, look for the line below, in `Deploy/config/nginx/nginx.conf` and remove it from the sections: `location /`, `location /v2/` and `location /service/`.
1.When setting up Harbor behind an nginx proxy or elastic load balancing, look for the line below, in `Deploy/config/nginx/nginx.conf` and remove it from the sections if the proxy already has similar settings: `location /`, `location /v2/` and `location /service/`.
```
proxy_set_header X-Forwarded-Proto $scheme;
```

10
docs/kubernetes_deployment.md
View File

@ -1,6 +1,6 @@
## Deploy harbor on kubernetes.
For now, it's a little tricky to start harbor on kubernetes because
1. Registry uses https, so we need cert or workaround to avoid errors like this:
## Deploying Harbor on Kubernetes
To deploy Harbor on Kubernetes, it requires some additional steps because
1. When Harbor registry uses https, so we need cert or workaround to avoid errors like this:
```
Error response from daemon: invalid registry endpoint https://{HOST}/v0/: unable to ping registry endpoint https://{HOST}/v0/
v2 ping attempt failed with error: Get https://{HOST}/v2/: EOF
@ -19,12 +19,12 @@ For now, it's a little tricky to start harbor on kubernetes because
sudo service docker restart
```
2. The registry config file need to know the IP (or DNS name) of the registry, but on kubernetes, you won't know the IP before the service is created. There are several workarounds to solve this problem for now:
2. The registry config file needs to have the IP (or DNS name) of the registry, but on Kubernetes, you don't know the IP before the service is created. There are several workarounds to solve this problem for now:
- Use DNS name and link th DNS name with the IP after the service is created.
- Rebuild the registry image with the service IP after the service is created and use ```kubectl rolling-update``` to update to the new image.
To start harbor on kubernetes, you first need to build the docker images. The docker images for deploying Harbor on Kubernetes depends on the docker images to deploy Harbor with docker-compose. So the first step is to build docker images with docker-compose. Before actually building the images, you need to first adjust the [configuration](https://github.com/vmware/harbor/blob/master/Deploy/harbor.cfg):
To start Harbor on Kubernetes, you first need to build the docker images. The docker images for deploying Harbor on Kubernetes depends on the docker images to deploy Harbor with docker-compose. So the first step is to build docker images with docker-compose. Before actually building the images, you need to first adjust the [configuration](https://github.com/vmware/harbor/blob/master/Deploy/harbor.cfg):
- Change the [hostname](https://github.com/vmware/harbor/blob/master/Deploy/harbor.cfg#L5) to ```localhost```
- Adjust the [email settings](https://github.com/vmware/harbor/blob/master/Deploy/harbor.cfg#L11) according to your needs.

8
docs/prepare-swagger.sh
View File

@ -9,10 +9,10 @@ wget https://github.com/swagger-api/swagger-ui/archive/v2.1.4.tar.gz -O swagger.
echo "Untarring Swagger UI package to the static file path..."
tar -C ../static/vendors -zxf swagger.tar.gz swagger-ui-2.1.4/dist
echo "Executing some processes..."
sed -i 's/http:\/\/petstore\.swagger\.io\/v2\/swagger\.json/'$SCHEME':\/\/'$SERVER_IP'\/static\/resources\/yaml\/swagger\.yaml/g' \
../static/vendors/swagger-ui-2.1.4/dist/index.html
sed -i.bak 's/http:\/\/petstore\.swagger\.io\/v2\/swagger\.json/'$SCHEME':\/\/'$SERVER_IP'\/static\/resources\/yaml\/swagger\.yaml/g' \
../static/vendors/swagger-ui-2.1.4/dist/index.html
mkdir -p ../static/resources/yaml
cp swagger.yaml ../static/resources/yaml
sed -i 's/host: localhost/host: '$SERVER_IP'/g' ../static/resources/yaml/swagger.yaml
sed -i 's/ \- http$/ \- '$SCHEME'/g' ../static/resources/yaml/swagger.yaml
sed -i.bak 's/host: localhost/host: '$SERVER_IP'/g' ../static/resources/yaml/swagger.yaml
sed -i.bak 's/ \- http$/ \- '$SCHEME'/g' ../static/resources/yaml/swagger.yaml
echo "Finish preparation for the Swagger UI."

19
migration/changelog.md Normal file
View File

@ -0,0 +1,19 @@
# What's New in Harbor Database Schema
Changelog for harbor database schema
## 0.1.0
## 0.1.1
- create table `project_member`
- create table `schema_version`
- drop table `user_project_role`
- drop table `project_role`
- add column `creation_time` to table `user`
- add column `sysadmin_flag` to table `user`
- add column `update_time` to table `user`
- add column `role_mask` to table `role`
- add column `update_time` to table `project`
- delete data `AMDRWS` from table `role`
- delete data `A` from table `access`

2
migration/migration_harbor/versions/0_1_1.py
View File

@ -83,7 +83,7 @@ def upgrade():
session.delete(role)
session.query(Role).update({Role.role_id: Role.role_id - 1})
#delete M from table access
#delete A from table access
acc = session.query(Access).filter_by(access_id=1).first()
session.delete(acc)
session.query(Access).update({Access.access_id: Access.access_id - 1})

2
models/project.go
View File

@ -34,4 +34,6 @@ type Project struct {
Togglable bool
UpdateTime time.Time `orm:"update_time" json:"update_time"`
Role int `json:"role_id"`
RepoCount int `json:"repo_count"`
}

1
ui/router.go
View File

@ -54,6 +54,7 @@ func initRouters() {
beego.Router("/api/search", &api.SearchAPI{})
beego.Router("/api/projects/:pid/members/?:mid", &api.ProjectMemberAPI{})
beego.Router("/api/projects/?:id", &api.ProjectAPI{})
beego.Router("/api/statistics", &api.StatisticAPI{})
beego.Router("/api/projects/:id/logs/filter", &api.ProjectAPI{}, "post:FilterAccessLog")
beego.Router("/api/users", &api.UserAPI{})
beego.Router("/api/users/?:id", &api.UserAPI{})