Closed OriHoch closed 7 years ago
this could be a python command that will run periodically via cron
this command should:
What's the base? py.test?
whatever you wish.. open to suggestions
@OriHoch Hi, I'm hasadna newbie. This issue should be pretty straightforward and i'd like to help.
Where the health check can be deployed? Does Hasadna have any budget for compute resources (AWS, Heroku, etc...)?
I've noticed that there's already code for the check. IS anything missing there?
Thanks, Benny
thanks, the metadata check code you linked to is a service that checks the metadata about the knesset services. not the data itself
this issue is about something similar that will check the data itself for any inconsistencies. at the most basic level it will check if the service returns http status 200 or not, more advanced stuff will be to open the data itself and use some kind of heuristics to determine if it's valid or not.
regarding deployment - we have everything needed to deploy this service. This issues should just be a python command which we will run somehow / somewhere
For fast solution, checking only the http status code, we can use a service like StatusCake, the free tier will be sufficient for this. Basically this is it: http://uptime.statuscake.com/?TestID=IEdmFQLe1V (I've created it on my free account, we should create an account for hasadna).
It has nice integrations, so when one of the endpoints is down you can trigger email, Slack, public tweet (could be nice!). If needed, it's possible to plug the email or slack to create a github issue. BTW, slack should be free for non profit orgs.
WDYT?
Sounds good
On Sun, Jun 5, 2016, 11:27 Benny Bauer notifications@github.com wrote:
For fast solution, checking only the http status code, we can use a service like StatusCake http://statuscake.com, the free tier will be sufficient for this. Basically this is it: http://uptime.statuscake.com/?TestID=IEdmFQLe1V (I've created it on my free account, we should create an account for hasadna).
It has nice integrations, so when one of the endpoints is down you can trigger email, Slack, public tweet (could be nice!). If needed, it's possible to plug the email or slack to create a github issue. BTW, slack should be free https://slack.com/pricing#slack_for_good for non profit orgs.
WDYT?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/hasadna/knesset-data/issues/4#issuecomment-223799853, or mute the thread https://github.com/notifications/unsubscribe/ABJLBgIu2ckjdczQg8hin3duIrz9A6ggks5qIohzgaJpZM4HRTpS .
So how do we take it from here?
Yes we have slack , I'll invite you
On Sun, Jun 5, 2016, 11:36 Benny Bauer notifications@github.com wrote:
So how do we take it from here?
- What should be the strategy for endpoint downtime handling? Open a github issue, or just sending a link to the dashboard? What should be the distribution list?
- Is there a slack team? do we want one?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/hasadna/knesset-data/issues/4#issuecomment-223800171, or mute the thread https://github.com/notifications/unsubscribe/ABJLBnoVNRChbKBbgUyQXwLUbIjPsTRqks5qIoqWgaJpZM4HRTpS .
@bennybauer - please send me you email, you can send me in twitter
@OriHoch and @bennybauer
So we can either save the tests for status cake in a repo and update via api https://www.statuscake.com/api/ Or, we keep the enterprise solutions for infrastructure monitoring, and provide an organization wide instance like Zabbix http://www.zabbix.com/license.php
personally I like the option of calling statuscake api from our server
@uda ,
If we're going on with statuscake please send me the credentials over slack and I'll handle this issue.
too much changed since this issue was opened, it's less relevant now that we use the datapackages
User Story
Open Knesset periodically (every night) gets votes data form The Knesset using the Knesset Dataservice.
On March. 2 there was a problem on Knesset side which caused votes api to stop working
Expected
Actual
Description
should have a process that periodically (daily) checks the knesset dataservice services and automatically open new issues if problems are encountered
it should check: