alerta / alerta-webui Goto Github PK
View Code? Open in Web Editor NEWAlerta Web UI 7.0
Home Page: https://try.alerta.io
License: Apache License 2.0
Alerta Web UI 7.0
Home Page: https://try.alerta.io
License: Apache License 2.0
Issue Summary
The orange bar always goes under All when back from an alert's detail view, even when it should be a more specific environment.
Environment
/config
endpoint output)To Reproduce
Steps to reproduce the behavior:
Expected behavior
The orange bar should be displayed under the 'CustomEnvironment'.
Additional context
I'm using custom ALLOWED_ENVIRONMENTS.
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Is your feature request related to a problem? Please describe.
We faced the situation where we had planned maintenance of the whole datacenter.
We needed to snooze alerts from the whole datacenter.
You can search for the alerts in Alerta Web UI, but multi-select works only one by one (and also only in Chrome) [win key] + [left mouse] which is useless if you have 500+ alerts you have to snooze.
Describe the solution you'd like
I would like to have the possibility to select a range of alerts, for example by selecting the first one with [win key] + [left mouse] and then the last one with [Shift] + [win key] + [left mouse].
Describe alternatives you've considered
Or alternatively to select all alerts on the actual screen where I limited my selection with Search/Filter.
Workaround
I have found a way to ack multiple alerts based on filter/query conditions using alerta cli.
# Search for all Open alerts from environments SIMU where event contains word deadman
alerta query -f status=open -f environment=SIMU -f event=~deadman
# Acknowledge them
alerta ack -f status=open -f environment=SIMU -f event=~deadman
After successful authentication with GitLab I stay on the login page. I expect that I should end up on the /alerts
page, for example, but all look like that redirect on the login page does not work. I always see the /login?redirect=%2Falerts
address in the address bar in my browser and the message Authenticating with GitLab ...
after authentication.
Environment
OS: Ubuntu 18.04
API version: 6.8.2
Deployment: self-hosted
WSGI environment: dev environment
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? GitLab
Customer views? No
web UI version: latest master
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
Authenticating with GitLab ...
messageExpected behavior
Alerta Beta should redirect me to the /alerts
page (or any previously opened page) after authentication.
Is your feature request related to a problem? Please describe.
My proposal does not solve any problem but makes using Alerta more convenient from my point of view.
Describe the solution you'd like
Now I cannot change the status comment without changing the current status of an alert but in some cases, it may be necessary. For instance, if the previous comment is not correct or the new one contains additional information. As a result, now I should use the following sequence: change status from A to B and then change the status back from B to A with a new comment - no so elegant.
I don't have any idea how it may be implemented with minimal efforts. Maybe, we may add a button like "Update comment" after "Add note" with the same look or something.
Issue Summary
The ui requests with every refresh the complete list of all alerts, no matter what is filtered. This in my test system already 2MB (gzipped 250kB) and vue is not doing a good job in handling all this data - Firefox is crashing after some minutes.
Environment
OS: Windows 7
API version: 7.1.0
Deployment: Cloud Foundry
For self-hosted, WSGI environment: nginx
Database: MongoDB
Server config:
Auth enabled? Yes
Auth provider? OpenID
Customer views? Yes
web UI version: 7.1.0
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
Expected behavior
The defined filter should be used for the /alerts
endpoint.
Issue Summary
Viewers in our system don't have any permissions except strictly required read:alerts and read:blackouts. But a webUI complains in every load:
"Missing required scope: read:customers (403)"
Is it really needed? Should it be documented at least?
Environment
To Reproduce
Steps to reproduce the behavior:
Alert appears in the bottom of the view: "Missing required scope: read:customers (403)"
Issue Summary
Every ops in our environment wants to watcxh only for open alerts. At the moment they have to configure filters every time they open Alerta UI to remove acked status from filter.
Environment
To Reproduce
Steps to reproduce the behavior:
Filter configuration is always the same showing open and acked alerts.
Expected behavior
A User should be a able to configure a default filter configuration.
Issue Summary
alexb145 @alexb145 13:54
@satterly this seems like a bug, we get this if we type anything in the search bar at the top then after searching click the x to delete it
this also makes all alerts disappear
and there's also things that we cannot find by searching using the top bar but which we can find using the filters on the right side
Environment
OS: [eg. Linux, Mac, Windows]
API version: [eg. 6.8.1]
Deployment: [eg. Docker, Heroku, AWS, GCloud, self-hosted]
For self-hosted, WSGI environment: [eg. nginx/uwsgi, apache/mod_wsgi]
Database: [eg. MongoDB, Postgres]
Server config:
Auth enabled? Yes/No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? Yes/No
(or provide the /config
endpoint output)
web UI version: [eg. 6.8.1]
CLI version: [eg. 6.8.1]
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
Additional context
Add any other context about the problem here.
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Issue Summary
If notification sounds are enabled, the sound should only play if there are Open alerts for the selected environment tab.
This does not seem to be the case. ie. if there are any Open alerts for any environment a notification sound will play even if the selected environment tab has no Open alerts.
Environment
OS: [eg. Linux, Mac, Windows]
API version: [eg. 6.8.1]
Deployment: [eg. Docker, Heroku, AWS, GCloud, self-hosted]
For self-hosted, WSGI environment: [eg. nginx/uwsgi, apache/mod_wsgi]
Database: [eg. MongoDB, Postgres]
Server config:
Auth enabled? Yes/No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? Yes/No
(or provide the /config
endpoint output)
web UI version: [eg. 6.8.1]
CLI version: [eg. 6.8.1]
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Additional context
None
Issue Summary
Invoke Google Analytics if trackingId not define.
Environment
OS: Linux
API version: alerta-6.8.1-py2.py3-none-any.whl alerta_server-7.2.1-py2.py3-none-any.whl
Deployment: Docker
Intranet without Internet access.
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? LDAP
Customer views? No
web UI version: 7.2.0
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Disable analytics plugins if trackingId not define
Issue Summary
Refreshing browser breaks alert details page
Environment
OS: Linux
API version: 7.2.0
Deployment: Docker
For self-hosted, WSGI environment: nginx/uwsgi
Database: MongoDB
Server config:
Auth enabled? No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? No
(or provide the /config
endpoint output)
web UI version: 7.2.0
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
Alert details should just redisplay.
Screenshots
above
Additional context
none
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Issue Summary
Google Chrome Out of memory
Environment
OS: Linux
API version: alerta-6.8.1-py2.py3-none-any.whl alerta_server-7.2.1-py2.py3-none-any.whl
Deployment: Docker
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? LDAP
Customer views? No
web UI version: 7.2.0
To Reproduce
Steps to reproduce the behavior:
Expected behavior
All work without errors.
Issue Summary
When I try to reload (or refresh, or press F5 or even Ctrl+F5) the page with alert details, I always see an empty page and a popup alert in my browser.
For example, I reload the page:http://192.168.1.25/alerta/alert/68f54758-f190-4557-8b84-bae6947518b4 and as a result, I see the following:
And after clicking on the OK button, I see an empty page.
In the web developer console, I see the following error:
Error: Network Error
at t.exports (createError.js:16)
at XMLHttpRequest.h.onerror (xhr.js:69)
localhost:8080/config:1 Failed to load resource: net::ERR_CONNECTION_REFUSED
createError.js:16 Uncaught (in promise) Error: Network Error
at t.exports (createError.js:16)
at XMLHttpRequest.h.onerror (xhr.js:69)
So Alerta WEB tries to use an incorrect Alerta endpoint (localhost:8080
instead 192.168.1.25
). And as I can see, it happens because Alerta UI loads config.json
from an incorrect place: http://192.168.1.25/alerta/alert/config.json instead http://192.168.1.25/alerta/config.json. So config.json
is loaded with a relative URL.
This approach with the relative URL leads to errors in some cases so using BASE_URL
or something should be more reliable.
Environment
OS: Linux
API version: 7.3.2
Deployment: self-hosted
WSGI environment: nginx/gunicorn
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? GitLab
web UI version: 7.3.2
CLI version: 7.3.0
To Reproduce
Steps to reproduce the behaviour:
Expected behavior
I expect no errors in UI
Additional context
To solve the problem I use the following Nginx location to load config.json
:
location ~ ^/alerta/(.+/)?config.json$ {
alias /etc/alerta/config.json;
}
Though this solution is working now it cannot be considered as a good solution for all cases.
EDIT: Please, pay attention to prefix /alerta
in my Nginx location. I use it because I run Alerta UI under the /alerta
path.
when i run this command as follow:
python3 -m http.server 8000
browse to url:
http://10.201.5.197:8000/
The following error occurred:
“ERROR: Failed to connect to Alerta API due to missing or invalid config.json file
Please confirm a config.json file exists,contains an "endpoint" setting and is in the same directory as the application index.html file”
config.json
{"endpoint": "http://localhost:8081"}
Issue Summary
Hey, can i know if the alerta docker is working fine? I tried to start up the docker image. I can access the api endpoint(http://localhost:8005/api/config) , but not the UI( http://localhost:8005). It returns me Error: Could not retrieve user preferences.
Environment
OS: [eg. Linux, Mac, Windows]
API version: [eg. 6.8.1]
Deployment: [eg. Docker, Heroku, AWS, GCloud, self-hosted]
For self-hosted, WSGI environment: [eg. nginx/uwsgi, apache/mod_wsgi]
Database: [eg. MongoDB, Postgres]
Server config:
Auth enabled? Yes/No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? Yes/No
(or provide the /config
endpoint output)
web UI version: [eg. 6.8.1]
CLI version: [eg. 6.8.1]
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
App should load and show the following ...
Would it be possible to change VueRouter from history to hash mode?
History mode makes deployment a little tricky: we need special catch-all rule and redirect all unknown locations to single index.html
.
I know that workarounds are described at https://router.vuejs.org/guide/essentials/history-mode.htm but it is not always possible when you do not have full access to the static web server configuration.
Is your feature request related to a problem? Please describe.
It's problematic at least for a new users to find out what icon initiates what action in the alert view (these small action icons in the right of every alert).
Describe the solution you'd like
I think these deserve to have a tooltips.
Alerta Web version : Latest
alerta 7.2.11
OS : Linux
Behind Nginx
We use grafana - prometheus Alert Manager to display count of alerts,we link alerta web ui url to grafana dashboard for showing critical or warning alerts
Earlier version of web UI supported this feature i.e when i click the count of critical alerts in grafana it redirects me to alerta filtered page to show only critical alerts for a particular service.
i upgraded web UI to 7.x and this is not working now.
Developer Logs Shows
Failed to load resource: the server responded with a status of 404 (Not Found)
Nginx Logs
[09/Jul/2019:18:27:26 -0700] "GET /alerts HTTP/1.1" 404 571 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/75.0.3770.100 Safari/537.36"
Nothing on the uwsgi logs.
I am using the Alerta docker image (v7.2.7) and it seems like changing the 'text' color in COLOR_MAP is not being reflected in the web ui.
Following configuration is passed for color_map:
COLOR_MAP = {
'severity': {
'critical': '#cccccc',
'major': '#e2e2e2',
'minor': '#ff9980',
'warning': '#f7f7f7',
'indeterminate': '#A6ACA8',
'cleared': '#00AA5A',
'normal': '#00AA5A',
'ok': '#00AA5A',
'information': '#00A1BC',
'debug': '#9D006D',
'security': '#333333',
'unknown': '#A6ACA8'
},
'text': 'white',
'highlight': 'skyblue'
}
severity colors are working fine but the text is always BLACK.
Issue Summary
deleting alert in details view should return you to alert summary list
Environment
OS: [eg. Linux, Mac, Windows]
API version: [eg. 6.8.1]
Deployment: [eg. Docker, Heroku, AWS, GCloud, self-hosted]
For self-hosted, WSGI environment: [eg. nginx/uwsgi, apache/mod_wsgi]
Database: [eg. MongoDB, Postgres]
Server config:
Auth enabled? Yes/No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? Yes/No
(or provide the /config
endpoint output)
web UI version: [eg. 6.8.1]
CLI version: [eg. 6.8.1]
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Additional context
Add any other context about the problem here.
Hi!
I do not see in the interface the ability to set access to specific objects ( groups,enviroments for example). Now I am faced with the choice of the system, apparently I like it, but I would like to set up strict access, I'm not sure that I can do it.
Issue Summary
alerts count is truncated and not visible. since badges are used .
Environment
OS: Linux
API version: 7.3.0
Deployment: Docker
Database: MongoDB
web UI version: 7.3.0
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
count is visible
Additional context
I'm not a vue/js experte but from quick reserach swithcing to us v-chip (https://vuetifyjs.com/en/components/chips) could be a possible solution
/**
Issue Summary
A clear and concise description of what the bug is.
Environment
OS: Linux/RHEL6
API version: [eg. 6.8.1]
Deployment:
One server is running Apache 2.4 and hosting alerta-webui with reverse proxy pointing to a second server, where alerta-server is running with UWSGI.
Database: MongoDB
Server config:
Auth enabled? No
Auth provider? N?A
Customer views? No
web UI version: master
CLI version: 7.2.10
To Reproduce
Steps to reproduce the behavior:
Expected behavior
I expect Alerta to reload properly when F5 button is hit. Instead, 404 error is generated. When I go to:
https://mysite.com/alerta/
URL is rewitten to:
https://mysite.com/alerta/alerts#status:open,ack;dateRange:,;sb:default;sd:1
After hitting F5, I get error 404 that /alerta/alerts does not exist.
I am upgrading from webui 3.0.0 where URL is different and refresh works:
https://mysite2.com/alerta/#/alerts?status=open&status=unknown
We plan to use notes for support engineers communication between shifts, but the current list of notes hidden too deep in interface. What about to show last note on the alert details page?
Issue Summary
A clear and concise description of what the bug is.
Environment
OS: [eg. Linux, Mac, Windows]
API version: [eg. 6.8.1]
Deployment: [eg. Docker, Heroku, AWS, GCloud, self-hosted]
For self-hosted, WSGI environment: [eg. nginx/uwsgi, apache/mod_wsgi]
Database: [eg. MongoDB, Postgres]
Server config:
Auth enabled? Yes/No
Auth provider? Google, GitHub, OpenID, Basic, LDAP, etc
Customer views? Yes/No
(or provide the /config
endpoint output)
web UI version: [eg. 6.8.1]
CLI version: [eg. 6.8.1]
To Reproduce
Steps to reproduce the behavior:
For web app issues, include any web browser JavaScript console errors.
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Additional context
Add any other context about the problem here.
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Is your feature request related to a problem? Please describe.
Not really a problem, just a suggestion regarding the hamburger button and Alerta tag on top left of page to make it more intuitive.
Describe the solution you'd like
When clicking the hamburger button, the animation currently redraws the hamburger button and Alerta tag. Instead I would suggest having both the hamburger button and Alerta tag static, having only the menu below appear, a bit like the desktop youtube webpage does it. Furthermore, having the Alerta tag be a button which brings you back to the main Alerts page would be of help.
Describe alternatives you've considered
Another option would be to move the hamburger button to the top right (next to the refresh button) and upon clicking have a popup menu go down.
Those are only suggestions, thanks for all your work with Alerta.
Issue Summary
The environment tabs at the top of the alerts list page do not get updated properly when new alerts appear in the list.
Environment
OS: Linux
API version: 7.2.5
Deployment: Docker
For self-hosted, WSGI environment: nginx/uwsgi
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? Keycloak
Customer views? No
(or provide the /config
endpoint output)
web UI version: 7.2.5
CLI version: 7.2.1
To Reproduce
Steps to reproduce the behavior:
Expected behavior
When the new alert appears in the alerts list page, the "Development" tab should also appear with a count of 1.
Additional context
Manually refreshing the page will display tabs properly:
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Is your feature request related to a problem? Please describe.
None of preconfigured date/time formats suits into our environment. The old web UI honored DATE_FORMAT_* settings in server, but the new one doesn't.
Describe the solution you'd like
Either support custom date/time format in UI or honor DATE_FORMAT_* configured in server.
After updating to the last master Alerta cannot load any necessary resources with 404 error.
I run Alerta WEB UI under the /alertabeta
path prefix and it worked fine before updating. But after updating Alerta cannot load any resources.
Also, Alerta does not load config.json
now and it is quite strange because this file contains some important settings:
{
"endpoint": "https://monitoring.example.com",
"base_path": "/alertabeta"
}
Is your feature request related to a problem? Please describe.
The old UI spreads columns quite evenly in the alerts view, but the new one aligns all content to the left and action buttons to the right. With hires quite wide displays it looks ... very weird.
Describe the solution you'd like
Spread columns more evenly based on the content of course.
Additional context
I will not attach huge screenshots, but if your screen is on smaller side, zoom out to look at hour the alerts view behaves in bigger hires screens.
There's a new option called tokenPath
which is a breaking change.
Issue Summary
Our alerts include links to relevant Grafana dashboards and Slack channels so that operators have a quick option to get additional alert context. Unfortunately the Alerta dashboard escapes the html links passed in, so that the entire HTML link is displayed rather than just the link text. Users see <a href="http://grafana.mycompany.com/dashboard/db/data-service-capacity">service dashboard</a>
instead of service dashboard
This should be configurable behavior, or perhaps just turned off.
Environment
OS: Linux
API version: [alerta-web 7.3.0]
Deployment: Docker
Database: MongoDB
Server config:
Auth enabled? No
Customer views?No
web UI version: [alerta-web 7.3.0]
CLI version: n/a
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Clickable links to grafana and slack on the front page of the alerta app
I see the 404 error in the DevTools console after authentication in Alerta Beta. It cannot load the /user/me/attributes
resource.
Environment
OS: Linux
API version: 6.8.2
Deployment: self-hosted
WSGI environment: dev environment
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? GitLab
Customer views? No
web UI version: latest master
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
/user/me/attributes:1 Failed to load resource: the server responded with a status of 404 (NOT FOUND)
vue-authenticate.es2015.js:459 Possible Unhandled Promise Rejection: Error: Request failed with status code 404
at t.exports (createError.js:16)
at t.exports (settle.js:18)
at XMLHttpRequest.f.(/alertabeta/anonymous function) (https://monitoring.sys.swiftserve.com/alertabeta/js/chunk-vendors.cca74b33.js:29:94124)
Expected behavior
No error/warning messages
Event's JSON data on the "data" tab displays in one line without any scroll bar. So I cannot inspect the whole raw data - only some fields of it. Would be better if I could see formatted JSON in multiple lines with indents and so on.
Environment
OS: Ubuntu 18.04
API version: 6.8.2
Deployment: self-hosted
WSGI environment: dev deployment
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? GitLab
Customer views? No
web UI version: latest master
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
Expected behavior
Possibility to see all raw data
Issue Summary
If a user's preference is dark theme, after a browser refresh the dark theme is not applied but instead defaults back to white theme. Only by going back to the settings page will the dark theme be applied again.
Environment
OS: Windows
API version: 7.2.0
Deployment: Docker
For self-hosted, WSGI environment: nginx/uwsgi
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? Basic
Customer views? No
(or provide the /config
endpoint output)
web UI version: 7.2.0
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
Expected behavior
After refreshing browser, dark theme should remain.
Screenshots
none
Additional context
Tried with both Firefox and Chrome.
NOTE: Please provide as much information about your issue as possible.
Failure to provide basic details about your specific environment make
it impossible to know if an issue has already been fixed, can delay a
response and may result in your issue being closed without a resolution.
Issue Summary
I would like to use the new UI but it is lagging more and more with every single event. When I have more than 20 events in the queue and move from one line to an other I have to wait up to one second til the highlighting is visible.
I've profiled the change of the status filter from ACK (3 events) to OPEN+ACK (82 events) in Firefox. This lasts many seconds in which the UI is not responding at all.
profile.zip
Environment
Expected behavior
The old UI performed much better without such hangs.
It is not so convenient when the state of the alerts table resetting after refreshing/reopening the related page. It even happens after returning to the alerts page from the details page. Under the table state I mean sort order, current filter setting and so on.
So it will be good if we save the table state in local storage and will use it after returning on the /alerts
page.
If I'm looking some information about an alert on the alert details page and then I press the refresh button then nothing happens. It is not so convenient because to see updated information about alert I must return to the alerts page, find there the alert (with playing with filter) and open the alert again.
Is it possible to refresh details and history tabs by clicking on the refresh button?
Issue Summary
Unable to open /alert/alert-id when coming from Reports. Everything else works.
Environment
OS: Linux 4.9.0-9-amd64 #1 SMP Debian 4.9.168-1+deb9u3 (2019-06-16) x86_64 GNU/Linux
Versions: API 7.3.0, Client 7.3.0, WebUI 7.3.0, Requests 2.22.0, Click 7.0, Python 3.5.3, Flask 1.1.0, Werkzeug 0.15.4
Deployment: Self-Hosted with Nginx and WSGI.
Database: MongoDB
Server config:
Auth enabled? Yes
Auth provider? Basic
Customer views? Yes
To Reproduce
Steps to reproduce the behavior:
Full Config Here
https://pastebin.com/KjUpnVSu
Expected behavior
Opens the alert history
All looks like I can filter alerts on the alerts page only by the Resource
and Description
columns. The same is true for searching alerts in the Search
box.
Is it possible to add filtering/searching by the Event
field?
Thank you for this improvement - it is really helpful and convenient now.
But I see two problems with the current approach:
Nonetheless, status comments work fine and I don't see any obvious problems with them right now. We may lose the user's comment when alert severity is changing, though, but I'm not sure about it and should test this behaviour.
What I see (pay attention to comment content):
Originally posted by @vykulakov in #234 (comment)
Issue Summary
I have built a bridge for pulling job states from a legacy component and pushing it into alerta, which means there are always around 100 alerts with several severities present across different environments. When the UI runs for a few hours without being interacted with, Firefox tells that the page is no longer responding. The UI itself is also pretty slow when moving the mouse over the alerts. A quick performance analysis using the integrated developer tool show that a major part of the time is spent in mouseenter/mouseexit/readystatechange event handlers and DOM updates.
Environment
OS: Ubuntu 16.04
API version: 7.4.0
Deployment: Docker
Database: Postgres
Server config:
{
"actions": [],
"alarm_model": {
"colors": {
"severity": {
"cleared": "#00CC00",
"critical": "red",
"debug": "#9D006D",
"indeterminate": "lightblue",
"informational": "#00CC00",
"major": "orange",
"minor": "yellow",
"normal": "#00CC00",
"ok": "#00CC00",
"security": "blue",
"trace": "#7554BF",
"unknown": "silver",
"warning": "dodgerblue"
},
"text": "black"
},
"defaults": {
"normal_severity": "normal",
"previous_severity": "indeterminate",
"status": "open"
},
"name": "Alerta 7.4.0",
"severity": {
"cleared": 7,
"critical": 1,
"debug": 8,
"indeterminate": 5,
"informational": 6,
"major": 2,
"minor": 3,
"normal": 7,
"ok": 7,
"security": 0,
"trace": 9,
"unknown": 10,
"warning": 4
},
"status": {
"ack": "C",
"assign": "B",
"blackout": "E",
"closed": "F",
"expired": "G",
"open": "A",
"shelved": "D",
"unknown": "H"
}
},
"audio": {
"new": null
},
"auth_required": true,
"aws_region": "us-east-1",
"azure_tenant": "common",
"client_id": "********",
"cognito_domain": null,
"colors": {
"severity": {
"cleared": "#00CC00",
"critical": "red",
"debug": "#9D006D",
"indeterminate": "lightblue",
"informational": "#00CC00",
"major": "orange",
"minor": "yellow",
"normal": "#00CC00",
"ok": "#00CC00",
"security": "blue",
"trace": "#7554BF",
"unknown": "silver",
"warning": "dodgerblue"
},
"text": "black"
},
"columns": [
"severity",
"status",
"lastReceiveTime",
"timeoutLeft",
"duplicateCount",
"customer",
"environment",
"service",
"resource",
"event",
"value"
],
"customer_views": false,
"dates": {
"longDate": "d/M/yyyy h:mm:ss.sss a",
"mediumDate": "EEE d MMM HH:mm",
"shortTime": "HH:mm"
},
"debug": false,
"email_verification": false,
"endpoint": "http://********/api",
"github_url": "https://github.com",
"gitlab_url": "https://********",
"indicators": {
"queries": [
{
"query": [
[
"environment",
"Production"
]
],
"text": "Production"
},
{
"query": [
[
"environment",
"Development"
]
],
"text": "Development"
},
{
"query": {
"q": "event:Heartbeat"
},
"text": "Heartbeats"
},
{
"query": "group=Misc",
"text": "Misc."
}
],
"severity": [
"critical",
"major",
"minor",
"warning",
"indeterminate",
"informational"
]
},
"keycloak_realm": null,
"keycloak_url": null,
"oidc_auth_url": "https://********/oauth/authorize",
"provider": "gitlab",
"refresh_interval": 5000,
"severity": {
"cleared": 7,
"critical": 1,
"debug": 8,
"indeterminate": 5,
"informational": 6,
"major": 2,
"minor": 3,
"normal": 7,
"ok": 7,
"security": 0,
"trace": 9,
"unknown": 10,
"warning": 4
},
"signup_enabled": false,
"site_logo_url": "",
"sort_by": "lastReceiveTime",
"tracking_id": null
}
To Reproduce
Expected behavior
No freezes, UX should be a little better.
On the main alerts page, would it be possible to add a filter for severity? Currently there is status, service, group, client (if applicable) and time.
Is your feature request related to a problem? Please describe.
While tracking down a bug where sometimes seemingly open alerts can't be acked (not yet there, probably combination of blackout, timeout, expire and my configuration), I noticed that API gives quite reasonable error messages with 409, but these are not shown to the user leaving a user clueless "why doesn't something happen if I click on some action"?
Describe the solution you'd like
Format InvalidAction() exceptions slightly better (correct case etc) and show these to the user if returned with 409 error.
Issue Summary
The alert list wastes space for the description column which isn't used in my case at all...
Environment
Expected behavior
Try to avoid wrapping lines when possible.
Issue Summary
Description is shown regardless of configuration. In our case it's show twice because COLUMNS is alertad.conf already contains 'text' element. But users might want to configure UI to hide description at all.
Environment
To Reproduce
Steps to reproduce the behavior:
Expected behavior
There shouldn't be two Description columns.
Alerta Beta always redirects me to the login page after refreshing any current page.
For instance, if I'm exploring alerts on the /alerts
page and then press the F5
button to refresh this page I always end up on the login page. It is quite strange because with the previous version of Alerta Web I always stay on the same page without any re-authentication (at least I do not see it).
Environment
OS: Ubuntu 18.04
API version: 6.8.2
Deployment: self-hosted
WSGI environment: dev environment
Database: Postgres
Server config:
Auth enabled? Yes
Auth provider? GitLab
Customer views? No
web UI version: latest master
CLI version: 6.8.1
To Reproduce
Steps to reproduce the behavior:
/alerts
Expected behavior
I should stay on the same page at the end of reloading without any re-authentication.
I use a single domain to access all components of my monitoring system. To make it possible I use NGINX to route requests for different locations to different destinations. For instance, I have the following configured locations:
https://mon.example.com/alerta/
https://mon.example.com/grafana/
https://mon.example.com/prometheus/
https://mon.example.com/alertmanager/
All components work fine except Alerta Beta (even old Alerta UI) because it uses absolute URLs in its routing engine.
I suppose that Alerta Beta must work behind a proxy with any URL prefix.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.