Seems like this is missing something as none of the api endpoints seem to exist.

/# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
d1525bedbd9c alerta/alerta-web:latest "/bin/sh -c '/config 3 minutes ago Up 30 minutes 0.0.0.0:80->80/tcp alerta-web
b40d673bc3ff mongo:latest "/entrypoint.sh mong 4 minutes ago Up 36 minutes 27017/tcp alerta-db

/# ENDPOINT=${1:-http://localhost:80}
/# curl -s -XPOST -H "Content-type: application/json" -H "Authorization: Key demo-key" ${ENDPOINT}/alert -d '
{
"resource": "host678:eth0",
"event": "HW:NIC:FAILED",
"group": "Hardware",
"severity": "major",
"environment": "Production",
"service": [
"Network"
],
"text": "Network interface eth0 is down.",
"value": "error"
}'

404 Not Found

Full list of server config environment variables should be here https://docs.alerta.io/en/latest/configuration.html#config-file-settings

See https://github.com/mytaxi/docker-alpine-alerta for example.

It's not possible to run the docker image behind a reverse proxy on a certain path, e.g. /alerta/. It's possible to set BASE_URL=/alerta/api but that only affects the API.

• The UI is still served from / - that can be workarounded by altering the path on the proxy to remove /alerta from the path
• Resources are loaded relative to the UI, so that's working
• The callback URL for oauth (e.g. for github) if always http(s)://mydomain.com/, so the path part is missing.

See https://github.com/alerta/docker-alerta/blob/master/alertad.conf.sh#L9

/cc @hrzbrg @bl1nk

Hi,
Is it possible to configure this Dockerfile to alpine environment. If yes can you please the link for it.

The docker-entrypoint.sh script dynamically creates a file "/etc/alertad.conf" on the first container run. This file is missing a plugins definition. On a subsequent restart the container will fail to start with the following in the logs:

++ echo
++ cut -d, -f1
+ ADMIN_USER=
++ echo mongodb://mongo:27017/monitoring
++ sed -e 's/mongodb:\/\///'
+ MONGO_ADDR=mongo:27017/monitoring
+ '[' '!' -f /app/config.js ']'
+ '[' '!' -f /etc/alertad.conf ']'
++ python -c 'exec(open('\''/etc/alertad.conf'\'')); print('\'','\''.join(PLUGINS))'
Traceback (most recent call last):
  File "<string>", line 1, in <module>
NameError: name 'PLUGINS' is not defined
+ PLUGINS=

On the second container startup, the following on line 33 causes the error:

PLUGINS=$(python -c "exec(open('$ALERTA_SVR_CONF_FILE')); print(','.join(PLUGINS))")

If you would like I could spend a few minutes and put in a pull request that fixes it (and still honors the PLUGINS environment variable if passed in from docker.)

https://github.com/alerta/docker-alerta/blob/master/docker-entrypoint.sh#L50

It should be possible for a container to use a user-supplied alerta.conf client config just as they can supply an alertad.conf file for server configs not supported by environment variables.

Hi,

I am trying to set up Gitlab auth and noticed that GITLAB_URL is not present in here.

I would expect the config to be something like this:

'use strict';
angular.module('config', [])
  .constant('config', {
    'endpoint'    : "${BASE_URL}",
    'provider'    : "${PROVIDER}",
    'gitlab_url'   : "${GITLAB_URL}"
    'client_id'   : "${OAUTH2_CLIENT_ID}",
    'colors'      : {}
  });

There seems to be a huge difference and some options are available in the kubernetes frontend config.
https://github.com/alerta/docker-alerta/blob/master/contrib/kubernetes/frontend/config.js.template

Are there any plans to support all these settings in the standard docker-entrypoint.sh

Hello,

Using the variables OAUTH2_CLIENT_SECRET and OAUTH2_CLIENT_ID to use gmail auth is not working.

Inside alertad.conf was ignored:

alerta@ae5a9a8df24b:/$ cat /app/alertad.conf 
DEBUG = True
AUTH_REQUIRED = False
OAUTH2_CLIENT_ID = 'xxxxxx'
OAUTH2_CLIENT_SECRET = 'yyyyyyyyyyyy'

Using docker-compose.yml was ignored also:

    environment:
      - DEBUG=1  # remove this line to turn DEBUG off
      - DATABASE_URL=postgres://postgres:postgres@db:5432/monitoring
      - AUTH_REQUIRED=False
      - OAUTH2_CLIENT_ID=xxxx
      - OAUTH2_CLIENT_SECRET=yyyy

My /login still the old one =/

Debug says that he's looking for my username in pgsql database, instead of google.com auth.

Thanks

As described in the tittle, it would be good to be able to add comment if the alert get shelved or closed so if someone goes and see why it's in that state you can check there faster than ask everyone in your team.
Also would be good to add simple comment in open alerts so if the alert still there and someone is working on it you can know that way

Hi Team,

do you know what is the best way to upgade alerta and keep exsting db? we are using it with docker compose and it says duplicate keys on admin when we try to upgrade

Cheers

With AUTH_REQUIRED=True set on the docker command line Alertmanager is unable to send alerts to Alerta (fails with HTTP 401 error code)

version: 6.5.0

When specifying plugins via INSTALL_PLUGINS I get the following error.

Installing plugin 'reject'
+ '[' -n '' ']'
+ IFS=,
+ for plugin in ${INSTALL_PLUGINS}
+ echo 'Installing plugin '\''reject'\'''
+ /venv/bin/pip install git+https://github.com/alerta/alerta-contrib.git#subdirectory=plugins/reject
The directory '/home/alerta/.cache/pip/http' or its parent directory is not owned by the current user and the cache has been disabled. Please check the permissions and owner of that directory. If executing pip with sudo, you may want sudo's -H flag.
The directory '/home/alerta/.cache/pip' or its parent directory is not owned by the current user and caching wheels has been disabled. check the permissions and owner of that directory. If executing pip with sudo, you may want sudo's -H flag.
Collecting git+https://github.com/alerta/alerta-contrib.git#subdirectory=plugins/reject
  Cloning https://github.com/alerta/alerta-contrib.git to /tmp/pip-req-build-edsodrg2
    Complete output from command python setup.py egg_info:
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "/venv/lib/python3.6/tokenize.py", line 452, in open
        buffer = _builtin_open(filename, 'rb')
    FileNotFoundError: [Errno 2] No such file or directory: '/tmp/pip-req-build-edsodrg2/plugins/reject/setup.py'
    
    ----------------------------------------

Cron is proved to be buggy in docker and can be removed from container with that simple hack in docker-compose.yaml

  alerta_cron_heartbeats:
    image: alerta/alerta-web:latest
    restart: always
    depends_on:
      - alerta-db
    volumes:
      - ./alerta/heartbeat.sh:/heartbeat.sh
    command: /heartbeat.sh

and

# cat ./alerta/heartbeat.sh
#!/bin/sh
sleep 60
alerta heartbeats --alert

it keeps all the things simplier.
container will be restarted by docker and looks like it is cron

• pre-install plugins as a layer
• Config file volumes
• What is alerta
• Environment variable info
• not linking db container
• Test “how to use this image” against entry point image
• Use labels?
• Alpine linux version?
• write tutorial for newbies

They seem to be ignored over OAUTH2_CLIENT_ID and OAUTH2_CLIENT_SECRET

Just noticed when trying to add plugins to the install images that the permissions still aren't quite right - its fine for running alertad - but the alerta user doesn't have access to install pip packages.

Will add a pull request to fix

Hi,

I have pulled the latest version of the docker-alerta repo and run the run.sh script (with my local customisations for email addresses and auth etc)

However whilst monitoring the logs on the mongo instance I see no connections being made to mongo.

I have tried this both on AWS and on centos 7.

Any pointers or further logs I can give you?

It seems that there's no supported way to add predefined key as apikey anywhere? Alerta generates keys automatically but i don't get how i can share them between multiple containers, or alternatively is there way to disable apikey authentication or use other kind of authentication on api calls?

Problem is that other container needs access to alerta api, however they don't know apikeys beforehand and creating them and configuring them afterwards outside (like ui and then configure them to containers) isn't really an option since same configuration is deployed multiple times by machine.

The plugin installer in the docker-entrypoint script fails for me, example:

I have no name!@3fa332413aa6:/$ /venv/bin/pip install git+https://github.com/alerta/alerta-contrib.git#subdirectory=plugins/prometheus
The directory '/.cache/pip/http' or its parent directory is not owned by the current user and the cache has been disabled. Please check the permissions and owner of that directory. If executing pip with sudo, you may want sudo's -H flag.
The directory '/.cache/pip' or its parent directory is not owned by the current user and caching wheels has been disabled. check the permissions and owner of that directory. If executing pip with sudo, you may want sudo's -H flag.
Collecting git+https://github.com/alerta/alerta-contrib.git#subdirectory=plugins/prometheus
  Cloning https://github.com/alerta/alerta-contrib.git to /tmp/pip-sttxqumy-build
fatal: unable to look up current user in the passwd file: no such user
shommand "git clone -q https://github.com/alerta/alerta-contrib.git /tmp/pip-sttxqumy-build" failed with error code 128 in None

I'll submit a PR for this which resolves the issue (unless I'm missing something and there's a better workaround?)

The new config.json and config.js and a few other bits that have changed between v5 and v6 cause the docker build in the contrib to fail.

angular-alerta-webui-master/web.js
mv: cannot stat '/usr/share/nginx/html/config.js': No such file or directory
The command '/bin/sh -c tar zxvf /tmp/web.tar.gz -C /tmp &&     rm -rf /usr/share/nginx/html &&     mv /tmp/angular-alerta-webui-master/app /usr/share/nginx/html &&     mv /usr/share/nginx/html/config.js /usr/share/nginx/html/config.js.orig' returned a non-zero code: 1
Makefile:10: recipe for target 'container' failed

probably could do with the same cleanup as the main alerta-web image

When requesting Prometheus metrics from Alerta running on Docker with

curl -XGET http://localhost:8080/management/metrics/

Nginx returns HTTP 404

Hi,

Im trying to deploy this container with a mongodb replica set for HA as advised by the documentation.
The docker-entrypoint script uses a mongodb client version 2.6 which needs a connection uri in the form:

<replset>/<host0>,<host1>...

If i set the MONGO_URI env var in such form to make the mongo client happy, alerta fails to connect to mongodb because of the invalid connection uri.

Setting MONGO_URI in alertad.conf doesn't help either because it seems the env var overwrites it.

PS: Thanks! Really awesome project 👍

Luis Felipe Dominguez Vega @lfdominguez 17:36
@satterly the solution is create the alertad.conf file and add the TELGRAM configurations, because in the environment variables dont work

with uwsgi try to log on /alertad.log

it's possible to send logs to stderr/stdout ?

2019-01-21 10:45:06,382 DEBG 'uwsgi' stdout output:
PermissionError: [Errno 13] Permission denied: '/alertad.log'

We're using Alerta 6.0.1 and I've noticed that the Ack button doesn't work for alerts that were sent from Prometheus Alertmanager even though the user has the admin role and scope and can execute all the other commands (Delete, Close, Open, Watch, Shelve). Sending similar alert with the alerta CLI produces and alert that can be acknowledged by the user. I'm wondering how this can happen. Was there a bug in this version, or if there is a problem with the way we've set up the admin user where it doesn't have permission to ack, but can do everything else. We plan to upgrade to Alerta 6.6 but don't have the time right now.

Hi,

I'm using the prometheus-alerta plugin - this lets me ack. alerts from alerta to alerta-manager.

Our alertmanager runs internally on SSL, so we need to provide the CA (I need to make the CA available inside this container).

I can get this working by having something like this in the entrypoint file:

# Install certs
if [ ! -z "$(ls -A /certs)" ]; then
  cp /certs/*.crt  /usr/local/share/ca-certificates/ 2>/dev/null
  update-ca-certificates
fi

This assumes that the user has mounted their certs into /certs/.

In addition to this we need to set the ca-bundle path for the python requests-lib.

Something like:
ENV REQUESTS_CA_BUNDLE=/etc/ssl/certs/ca-certificates.crt

The problem I have is that ca-update-certificates needs to run as root and the container-user is alerta.

I've tried a few workarounds (i.e. run the container as root, but su - alerta to run the rest of the entrypoint), but these end up rather messy (with multiple permissions issues).

Although I have it working in my fork, I'm running everything as root (which I don't think is right, so don't want to create a PR yet).

Do you have any advice? Not sure if this is something you have dealt with before on this project.

Thanks

I've custom built a docker container containing the mailer integration:
alerta-mailer-DockerFile

FROM alerta/alerta-web

USER root
RUN apt-get update
RUN apt-get -y upgrade
RUN apt-get autoclean
RUN apt-get clean
USER 1001

RUN /venv/bin/pip install --upgrade pip
RUN /venv/bin/pip install git+https://github.com/alerta/alerta-contrib.git#subdirectory=integrations/mailer

I've spun up the new image in a container and everything works from the web gui.
I exec into the alerta container and run: alerta-mailer
I see the following output:

DEBUG:mailer:Looking for rules files in /app/alerta.rules.d
Python dns.resolver unavailable. The skip_mta option will be forced to FalseError 99 connecting to localhost:6379. Cannot assign requested address.
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): localhost:8080
Starting new HTTP connection (1): localhost:8080
DEBUG:urllib3.connectionpool:http://localhost:8080 "POST /heartbeat HTTP/1.1" 404 None
http://localhost:8080 "POST /heartbeat HTTP/1.1" 404 None
DEBUG:urllib3.connectionpool:http://localhost:8080 "POST /heartbeat HTTP/1.1" 404 None
http://localhost:8080 "POST /heartbeat HTTP/1.1" 404 None

This looks like it's trying to post to the heartbeats page in alerta but failing as the path is incorrect.
I've run:
curl http://localhost:8080 and can confirm the page is returned. However, running http://localhost:8080/heartbeat fails with a 404 error which I expect is what I'm seeing from alerta-mailer.
curl -X POST http://localhost:8080/heartbeat

I get a 404 Not Found.

I'm building a helm chart for alerta, and I'd like to know how the CORS_ORIGINS are set. It's currently an env var, but I wanna know if it's comma separated or something, so I know if I can loop through it

Hi!

Is it my idea, or alertad by default is trying to reach mongo from 'localhost', those linking to a mongo container doesn't actually do anything usefull?

Thanks!!

Using the example docker-compose.yml the coker-compose up` fails with following error message

ERROR: manifest for alerta/alerta-web:latest not found

Checking on https://hub.docker.com/r/alerta/alerta-web/tags/ shows there is no image with tag latest. If the docker-compose.yml is amended to

  ...
  web:
    image: alerta/alerta-web:5.2.4
    ports:
  ...

the example can be executed.

Not sure if this is either to be changed on docker-alerta or on alerta-web to provide a docker image with the tag latest.

Alerta 5.x supports PostgreSQL so update the docker entry point script

Warning: The --link flag is a deprecated legacy feature of Docker. It may eventually be removed. Unless you absolutely need to continue using it, we recommend that you use user-defined networks to facilitate communication between two containers instead of using --link. One feature that user-defined networks do not support that you can do with --link is sharing environmental variables between containers. However, you can use other mechanisms such as volumes to share environment variables between containers in a more controlled way.

https://docs.docker.com/engine/userguide/networking/default_network/dockerlinks/

Add housekeeping job for heartbeats.

Hi,

I'm attempting to create a routing plugin within the docker version of alerta. With that said I've customized the Dockerfile from the repo a bit because I don't want to run the web console on the same container as my API container.

I'm basically following the tutorial here and the following is in routing.py:

def rules(alert, plugins):
    if alert.duplicateCount > 2:
        return [plugins['slack']]
    else:
        return []

The setup.py file is basically identical to the tutorial. I exec into the API server container, create routing.py using the above and setup.py. I activate the venv and install the plugin:

alerta@33958ff7cd34:/$ cd /app/plugins/routing/
alerta@33958ff7cd34:/app/plugins/routing$ . /venv/bin/activate
(venv) alerta@33958ff7cd34:/app/plugins/routing$ python setup.py install
running install
running bdist_egg
running egg_info
writing alerta_routing.egg-info/PKG-INFO
...

I restart the API server container, and when I try to send some test alerts with the alerta CLI I'm always seeing the following in the logs:

2018-08-26 19:21:05,083 DEBG 'uwsgi' stdout output:
2018-08-26 19:21:05,083 - alerta.plugins[30]: WARNING - Plugin routing rules failed: 'Alert' object has no attribute 'duplicateCount' [in /venv/lib/python3.6/site-packages/alerta/utils/plugin.py:45]

Any idea what the issue might be? I tried changing routing.py to use alert.duplicate_count (per the Alert model source) but this didn't seem to work either.

Hi,

I've set the authentication with Github and it's working with all the users in my organisations. However, none of the Github Users have admin access. We can manage users only in Basic Auth, so I can't change scopes and permissions...
Is there a solution to set all the github's users as admin, or a workaround to be able to log as an admin ?

Hi,

It looks like mailer fails to install as it doesn't reside in the "plugins" directory in the project.
The docker-entrypoint.sh script in alerta runs the following:

# Install plugins
  IFS=","
  for plugin in ${INSTALL_PLUGINS}
  do
    echo "Installing plugin '${plugin}'"
    /venv/bin/pip install git+https://github.com/alerta/alerta-contrib.git#subdirectory=plugins/$plugin
  done
  touch ${RUN_ONCE}
fi

This fails to install mailer (I'm assuming of course that I'm supposed to define it under plugins as there doesn't appear to be anywhere else I can define it in the alertad.conf file and very little direction as to how to integrate it with alerta).

To get around this, I can run:

/venv/bin/pip install git+https://github.com/alerta/alerta-contrib.git#subdirectory=integrations/mailer
Building wheels for collected packages: alerta-mailer
  Building wheel for alerta-mailer (setup.py) ... done
  Stored in directory: /tmp/pip-ephem-wheel-cache-2e1nah9o/wheels/9a/49/c1/b19cd89b13b7dc6802bbe04dd0edb93d0a23dcc54948c2c916
Successfully built alerta-mailer
Installing collected packages: vine, amqp, kombu, redis, alerta-mailer
Successfully installed alerta-mailer-5.2.0 amqp-2.4.1 kombu-4.3.0 redis-3.1.0 vine-1.2.0

from within the container but obviously not ideal. Additional to that is that despite it saying the alerta-mailer-5.2.0 package has been built and dropped into /tmp/pip-ephem-wheel-cache-2e1nah9o/wheels/9a/49/c1/b19cd89b13b7dc6802bbe04dd0edb93d0a23dcc54948c2c916, that directory does not exist. Even if it did, there is no clear instructions on how to make that implement with alerta.

If anyone can point out if I'm missing something, I'd appreciate it.

Hi,

I'm running Alerta in a docker container on DC/OS platform. Everytime I restart Alerta, it creates a new container, so I have to save the configuration in volumes to keep it across reboots. Currently, the API keys generated via the WebUI are lost. Can you tell me which files I need to put in volumes to make the configuration persistent ?

I have just configured nagios - alerta integration. Following the steps in https://github.com/alerta/nagios-alerta and I'm not recieving any alert or any error log either in Alerta or Nagios

When running the API server in the docker container the web UI works however when I got to access the API link I get the following...

{"code":500,"errors":["Traceback (most recent call last):\n File "/venv/lib/python3.6/site-packages/werkzeug/routing.py", line 1538, in match\n rv = rule.match(path, method)\n File "/venv/lib/python3.6/site-packages/werkzeug/routing.py", line 776, in match\n raise RequestSlash()\nwerkzeug.routing.RequestSlash\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n File "/venv/lib/python3.6/site-packages/flask/app.py", line 1813, in full_dispatch_request\n rv = self.dispatch_request()\n File "/venv/lib/python3.6/site-packages/flask/app.py", line 1791, in dispatch_request\n self.raise_routing_exception(req)\n File "/venv/lib/python3.6/site-packages/flask/app.py", line 1774, in raise_routing_exception\n raise request.routing_exception\n File "/venv/lib/python3.6/site-packages/flask/ctx.py", line 336, in match_request\n self.url_adapter.match(return_rule=True)\n File "/venv/lib/python3.6/site-packages/werkzeug/routing.py", line 1542, in match\n safe='/:|+') + '/', query_args))\nwerkzeug.routing.RequestRedirect: 301 Moved Permanently: None\n"],"message":"301 Moved Permanently: None","status":"error"}

Is there a way to fix this? I followed the directions for the docker container to the letter.

I get a 403 http response while using the docker alerta container and the nagios integration when I try to set an environment. It looks like I have to set an ALLOWED_ENVIRONMENT in alertad.conf? But there doesn't appear to be an environment variable to set that at docker run time.

Basic LDAP integration not working and not logging errors. Also alerta login.html not prepared to support username in login page without email

Some problems faced

1. AMQP

error: Unknown option auto_start_request
error: Unknown option use_greenlets

I had to review lines of the file usr/local/lib/python2.7/dist-packages/kombu/transport/mongodb.py:

            # 'auto_start_request': True,
            # auto_start_request = options ['auto_start_request'],

            # use_greenlets = _detect_environment ()! = 'default'

For these option is deprecated

2. Twilio

I entered the container shell, cloned the repo contrib, performed 'python setup' for Twilio, and had to change the file:

/usr/local/lib/python2.7/dist-packages/alerta_twilio-0.1.0-py2.7.egg/twilio_sms.py

with my credentials!

It did not work, export the variables.

Then I needed to enable the alertad.conf.sh:

PLUGINS = ['AMQP', 'reject', 'twilio_sms']

AMQP_URL = 'mongodb: // localhost: 27017 / kombu'
AMQP_QUEUE = 'alerts'
AMQP_TOPIC = 'notify'

All changes always with stop / start container!

I just wanted to share it if someone tb go through it!

Describe the bug
Running any of the command line options when exec'ed into a docker container fails with the same error messages.

To Reproduce
Steps to reproduce the behavior:

1. sudo docker exec -it alerta /bin/bash
2. Run 'alerta watch' or 'alerta top' or any other switch for alerta and the same error messages will be returned:

alerta@071c919c2588:/$ alerta watch
Traceback (most recent call last):
  File "/venv/bin/alerta", line 10, in <module>
    sys.exit(cli())
  File "/venv/lib/python3.6/site-packages/click/core.py", line 764, in __call__
    return self.main(*args, **kwargs)
  File "/venv/lib/python3.6/site-packages/click/core.py", line 717, in main
    rv = self.invoke(ctx)
  File "/venv/lib/python3.6/site-packages/click/core.py", line 1134, in invoke
    Command.invoke(self, ctx)
  File "/venv/lib/python3.6/site-packages/click/core.py", line 956, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/venv/lib/python3.6/site-packages/click/core.py", line 555, in invoke
    return callback(*args, **kwargs)
  File "/venv/lib/python3.6/site-packages/click/decorators.py", line 17, in new_func
    return f(get_current_context(), *args, **kwargs)
  File "/venv/lib/python3.6/site-packages/alertaclient/cli.py", line 54, in cli
    config.get_remote_config()
  File "/venv/lib/python3.6/site-packages/alertaclient/config.py", line 57, in get_remote_config
    remote_config = r.json()
  File "/venv/lib/python3.6/site-packages/requests/models.py", line 897, in json
    return complexjson.loads(self.text, **kwargs)
  File "/usr/local/lib/python3.6/json/__init__.py", line 354, in loads
    return _default_decoder.decode(s)
  File "/usr/local/lib/python3.6/json/decoder.py", line 339, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/usr/local/lib/python3.6/json/decoder.py", line 357, in raw_decode
    raise JSONDecodeError("Expecting value", s, err.value) from None
json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)

Expected behavior
A terminal view of the alerts or some other data back such as Uptime, etc.

Desktop:

• OS: Ubuntu 18.04 - Docker 18.9.2

Docker image alerta/alerta-web:5.2.4 run alerta-server 5.2.3
See below the result of pip freeze in alerta-web:5.2.4 container

docker run -ti --rm --name alerta --entrypoint "bash" alerta/alerta-web:5.2.4
alerta@1b466c311dae:/$ source venv/bin/activate
(venv) alerta@1b466c311dae:/$ pip freeze | grep "alerta"
alerta==5.2.0
alerta-server==5.2.3

is this expected behaviour ?

Labels with the Dockerfile ? eg. LABEL version="5.0.8"
Or tags when building on Docker hub?

as soon as I run the following command : docker run --name alerta-web --link alerta-db:mongo -d -p 80:80 alerta/alerta-web
I see container is in exited status like below.
[root@master docker-alerta]# docker ps -a
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
579634042fd3 alerta/alerta-web "/docker-entrypoin..." 5 seconds ago Exited (1) 4 seconds ago alerta-web
b19bf630f240 mongo "docker-entrypoint..." 39 minutes ago Up 39 minutes 27017/tcp alerta-db

What can be the fix for this ?