Hi,

Today, I had a meeting with Michaël and had discussions about work items
from the last week which basically includes:
- Connect Sugarizer-School-Portal with helm and add a feature to install a
simple chart from the interface.
- Add more translations for the Deployments part and update the error codes.
- Fix bugs and add improvements in current Sugarizer-School-Portal

Problems experienced last week: None

Here is a list of tasks that we have decided for this week:
- Connect local-sugarizer-chart with SSP.
- Add feature to add create an admin in Sugarizer-Server deployment through
SSP.
- Look into implementing the DNS feature in SSP.

PFA slack logs for reference. Next meeting is scheduled for next Monday.

Regards,
Nikhil
Michaël Ohayon  3:03 PM
Hello Nikhil :wave:

Nikhil  3:04 PM
Hello @Michaël Ohayon
3:04
How are you doing?

Michaël Ohayon  3:05 PM
Great and you? :smile:

Nikhil  3:05 PM
Well corona is spreading in India at an alarming rate these days. However, I'm in a safe zone.

Michaël Ohayon  3:06 PM
Glad you’re fine, hope you and your family won’t get touched.

Nikhil  3:06 PM
Yeah. Hope so
3:07
So the last week tasks were:
Connect Sugarizer-School-Portal with helm and add feature to install simple chart from the interface.
Add more translations for the Deployments part and update the error codes.
Fix bugs and add improvements in current Sugarizer-School-Portal
3:07
I have improved the Search feature in the SSP.
3:08
On clicking on Deploy, the SSP installs the Redis Helm chart in the kube cluster.
3:08
On clicking on Stops, it uninstalls the chart.
3:09
I've also improved the home page by adding some cards and tables.
3:10
Have you tested the SSP? I added few changes last night.

Michaël Ohayon  3:11 PM
I pulled the lastest modifications 1h ago
3:11
I forgot to add the bitnami repo (edited) 

Nikhil  3:11 PM
Okay

Michaël Ohayon  3:11 PM
(I did a new fresh install)
3:11
It worked fine on my latest update

Nikhil  3:12 PM
I've modified the readme to include instruction about MongoDB setup and I've also added [CHANGELOG.md](https://github.com/NikhilM98/sugarizer-school-portal-server/blob/master/CHANGELOG.md) to keep track of major changes

Michaël Ohayon  3:12 PM
Delete works perfectly :ok_hand:
:+1:
1

3:12
That’s very nice congrats
3:12
And the fact you have to disable the deployment first is a good security
:+1:
1


Nikhil  3:12 PM
Thanks :grinning:

Michaël Ohayon  3:13 PM
Latest thing I did not try yet is search

Nikhil  3:14 PM
Okay

Michaël Ohayon  3:14 PM
It’s working, I will try it further more with some funny things like accents é à (edited) 
:open_mouth:
1


Nikhil  3:14 PM
Have to seen the Home view? On clicking on the entries on the table, you can view the deployment.

Michaël Ohayon  3:16 PM
Yes that is very nice, I really can’t wait to fill it with high numbers :smile:

Nikhil  3:16 PM
Also, the role based restriction restricts the Moderator to only view the deployments

Michaël Ohayon  3:16 PM
There are some small texts missing translations but that is also working nice (Total Requests for instance)

Nikhil  3:16 PM
And clients can view/edit only their own deployments.

Michaël Ohayon  3:16 PM
Ah nice, I will try to test this as well

Nikhil  3:17 PM
Yeah
3:19
GSoC coding period officially begins from today :joy:

Michaël Ohayon  3:20 PM
Yes that is very nice to have all this already setup :joy:
3:20
I’m making some repetition here but thanks for your work that is very nice

Nikhil  3:21 PM
Thanks
3:21
So what should we do this week?
Should we actually connect the local-sugarizer-chart with the SSP?

Michaël Ohayon  3:22 PM
That would be nice
3:22
So we’ll get something up for testing

Nikhil  3:23 PM
When I asked you about cloud-dns, you told me this:
Maybe we will attach *.domain to the cluster endpoints ?
Best way would be some https://knative.dev/docs/serving/using-external-dns-on-gcp/#set-up-kubernetes-engine-cluster-with-clouddns-readwrite-permissions
Additionnal link: https://github.com/kubernetes-sigs/external-dns
Maybe I can take a deeper look in this.

Michaël Ohayon  3:24 PM
Yes that would be a necessary thing I guess
3:24
:smile:

Nikhil  3:25 PM
I was thinking of adding a feature in SSP to create Admin users in the Sugarizer-Server deployment. The person who requested the deployment or the admin would be able to add Admin user in the Sugarizer-Server deployment.
3:26
We also have to work on creating MongoDB backup and restore strategy. (edited) 

Michaël Ohayon  3:28 PM
First step would be to have a basic connection to a mongodb helm deployed somewhere
3:28
even on the same cluster I guess

Nikhil  3:32 PM
I found some blogs on MongoDB backup strategy during the research period of the project. I think that the point that our cluster will be on GKE will make a difference here. I think we can wait until we actually deploy the project on GKE before discussing the backup/restore strategy in detail.

Michaël Ohayon  3:32 PM
I’m not sure how much it will change to be on GKE vs on local for this purpose :thinking_face:
3:32
But that’s not the priority yes
3:33
Let’s keep on access with dns, “real platform” with sugarizer chart :smile:
3:33
And we’ll see afterwards
3:34
At some point we will have to deploy the ssp on the cluster, I was thinking that an Helm chart would be the best

Nikhil  3:34 PM
GKE uses different kind of persistent volumes. That's why I thought that it might make a difference. Still not sure.

Michaël Ohayon  3:34 PM
After we could improve with ansible
3:34
Ah yes, we could backup the volume. I’m not sure how safe it would be if there is someone writing on it
3:34
But yes that will changes things to be based on volumes

Nikhil  3:34 PM
Yeah. Right

Michaël Ohayon  3:35 PM
I would imagine a worker going at some acceptable rate
3:35
But that’s another nice subject to think of :smile:

Nikhil  3:36 PM
I would imagine a worker going at some acceptable rate
I didn't understand this point. Are workers like cron jobs in GCP?

Michaël Ohayon  3:36 PM
There’s nothing built it I guess
3:36
I’m not sure how to orchestrate things like that :smile:

Nikhil  3:36 PM
Okay
Pinned by you
3:38
Okay, so moving back to the tasks for this week:
Connect local-sugarizer-chart with SSP.
Add feature to add create admin through SSP.
Look into the implementing DNS feature in SSP.

Michaël Ohayon  3:38 PM
Did you have any issues during the past week?

Nikhil  3:38 PM
No. Not at all

Michaël Ohayon  3:38 PM
:smile:

Nikhil  3:39 PM
You mentioned that your company can provide me with GCP credentials to test SSP. Will it be possible?

Michaël Ohayon  3:41 PM
By the company I can give you admin token  on the cluster but I’ve not managed a way to give you access to the google dashboard without leaking credentials
3:41
So I think I’ll see with @Lionel how we can fund that

Nikhil  3:41 PM
Okay

Michaël Ohayon  3:41 PM
I can still setup the cluster if you want to deploy using helm on it
3:42
But that won’t help with disk strategy and stuff :disappointed: (edited) 

Nikhil  3:44 PM
I just wanted to check that what kind of volume storage changes we'll need to make and what kind of changes will work with cloud DNS. We can wait at least 2/3 weeks before we actually test it in practice.
3:45
I think 2/3 weeks would be enough to find some way to provide some way to authenticate, or maybe you yourself can test if that what actually works or not.
3:47
I think next week we can focus on certificate creation, DNS and other networking related issues.
3:49
After that I was thinking maybe we can test if the things actually work or not. Also, it'll be helpful in preparing an ansible package -- because we'll know what issues we face while deployment.

Michaël Ohayon  3:49 PM
Ok so I’ve got a week to figure out how to get credentials :smile:

Nikhil  3:50 PM
I think 2-3 weeks. There's no hurry, we can do something else meanwhile.

Michaël Ohayon  3:50 PM
For certificate it may work locally but I guess you’ll run quickly into issues

Nikhil  3:51 PM
Yeah
:+1:
1


Michaël Ohayon  3:54 PM
Do you have any more questions?

Nikhil  3:54 PM
No. Not for now

Michaël Ohayon  3:56 PM
Could you tell me if you have access to https://console.cloud.google.com/home/dashboard?project=mohayon-experiments ?
accounts.google.comaccounts.google.com
Google Cloud Platform
Google Cloud Platform lets you build, deploy, and scale applications, websites, and services on the same infrastructure as Google.

Nikhil  3:57 PM
Which account did you give access to?
3:57
nikhilmehra...@gmail.com
or
nme...@ph.iitr.ac.in
?

Michaël Ohayon  3:57 PM
the gmail one

Nikhil  3:57 PM
Okay
3:58
Yeah. I've access to the dashboard

Michaël Ohayon  3:58 PM
Ok I will try to use that to give you access to the required apis

Nikhil  3:58 PM
Nice
4:02
Should we end this meeting?

Michaël Ohayon  4:02 PM
Yes that is ok I guess
_______________________________________________
Sugar-devel mailing list
Sugar-devel@lists.sugarlabs.org
http://lists.sugarlabs.org/listinfo/sugar-devel

Reply via email to