Hi Giannis, Awaiting your reply to the draft proposal I sent in my previous email. Please review and let me know how I can develop this further. Thanks, Sanskar On Tue, Apr 6, 2021 at 12:48 AM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> wrote: > Hi Giannis > I have submitted the draft proposal taking into account the feedback you > gave , requesting again to have a look at that and provide your comments. > Proposal Link: Link > <https://docs.google.com/document/d/1Y8QcNpfEE6hS5SDUAiXsWQ37TuXsju0pzdGmJYU6vAM/edit#heading=h.48wu77618rar> > > Thanks and Regards > Sanskar Gupta > > On Mon, Apr 5, 2021 at 6:02 PM Giannis E. Skitsas <iskitsas [ at ] gmail [ dot ] com> > wrote: > >> Hi Sanskar, >> >> A user who has edit permissions should be able to access lock endpoints. >> >> Cached endpoints are usually accessible by all roles, probably from users >> having been authenticated. It depends on how someone wants to configure his >> setup. For example he might expose resources from CMS in the public, >> whereas others to be protected. >> >> Regards, >> Giannis >> -- >> Ioannis Skitsas >> >> >> >> On Mon, Apr 5, 2021 at 6:23 PM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> wrote: >> >>> Hi Giannis >>> Can you tell me which roles would be able to access the lock and cache >>> endpoints? >>> >>> Thanks >>> Sanskar Gupta >>> >>> On Sun, Apr 4, 2021 at 1:27 AM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> wrote: >>> >>>> Hi Giannis >>>> Yeah totally you are right, connecting backend and frontend should be >>>> pretty straight forward using kubernetes services ,wanted to know your >>>> thoughts if we can follow the same approach to connect prometheus to >>>> frontend >>>> >>>> Thanks and Regards >>>> Sanskar >>>> >>>> On Fri, Apr 2, 2021 at 11:05 PM Giannis E. Skitsas <iskitsas [ at ] gmail [ dot ] com> >>>> wrote: >>>> >>>>> Hi Sanskar, >>>>> >>>>> It could be useful for your dockerized service. Exposing on a >>>>> different port is another solution. It is up to you to decide. >>>>> >>>>> On the other hand, if you deploy to kubernetes you might not need >>>>> nginx. >>>>> >>>>> Do you have something specific in mind? Eg accessing prometheus >>>>> metrics from react frontend and drawing graphs? >>>>> >>>>> Regards, >>>>> Giannis >>>>> >>>>> On Fri, 2 Apr 2021 at 03:57, SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> wrote: >>>>> >>>>>> Hi Giannis >>>>>> I have a doubt: >>>>>> As frontend will be a separate deployment , we can use nginx to proxy >>>>>> request to cms-backend, can we do the same for HTTP API of prometheus as >>>>>> well using the same nginx.conf ? >>>>>> >>>>>> >>>>>> Thanks and regards >>>>>> Sanskar Gupta >>>>>> >>>>>> On Thu, Apr 1, 2021 at 12:25 PM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> >>>>>> wrote: >>>>>> >>>>>>> Hi Giannis >>>>>>> Thanks a lot for the valuable suggestions :) , will modify it right >>>>>>> away. >>>>>>> >>>>>>> Thanks and regards >>>>>>> Sanskar Gupta >>>>>>> >>>>>>> On Thu, Apr 1, 2021 at 12:18 PM Giannis E. Skitsas < >>>>>>> iskitsas [ at ] gmail [ dot ] com> wrote: >>>>>>> >>>>>>>> Hi Sanskar, >>>>>>>> >>>>>>>> Have a look at this proposal template >>>>>>>> <https://ellak.gr/wiki/index.php?title=Proposal_Template> for >>>>>>>> general instructions. You can also find others over the internet and >>>>>>>> combine suggestions. >>>>>>>> >>>>>>>> For your proposal I have the following comments: >>>>>>>> >>>>>>>> - It should be clear the project you are referring to by the >>>>>>>> official title and a link to the project. >>>>>>>> - In your project details section I would suggest to be more >>>>>>>> specific and clear. You can provide a better description based on the info >>>>>>>> you already have >>>>>>>> - I suggest using better headlines on your "Problem Statement" >>>>>>>> section. If you were about to create a table of contents how would those >>>>>>>> headlines (P1,P2) would look? >>>>>>>> - You can provide a section summarising all the technologies >>>>>>>> you will be using. It would be easier for the reader to instantly identify >>>>>>>> the technology stack without having to read all the sections >>>>>>>> - I suggest to re-organize your sections (mostly in project >>>>>>>> details) related to any new architecture or implementation, so it is more >>>>>>>> clear what you will be doing. (Sub)Sections/headlines proposals could be >>>>>>>> something like: Frontend Implementation , Monitoring and alerts, >>>>>>>> Performance improvement, Extending backend features, User management, >>>>>>>> Security etc >>>>>>>> >>>>>>>> Thanks, >>>>>>>> Giannis >>>>>>>> -- >>>>>>>> Ioannis Skitsas >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Wed, Mar 31, 2021 at 4:18 PM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Hi Giannis >>>>>>>>> I have prepared a proposal for GSoC 2021 [CMS Project] where I >>>>>>>>> tried to explain the ideas we discussed and how I plan to implement them , >>>>>>>>> request you to >>>>>>>>> have a quick look at that and provide me with some feedback. >>>>>>>>> PFA >>>>>>>>> >>>>>>>>> Thanks and Regards >>>>>>>>> Sanskar Gupta >>>>>>>>> >>>>>>>>> >>>>>>>>> On Mon, Mar 29, 2021 at 8:47 AM Giannis E. Skitsas < >>>>>>>>> iskitsas [ at ] gmail [ dot ] com> wrote: >>>>>>>>> >>>>>>>>>> Hi Sanskar, >>>>>>>>>> >>>>>>>>>> The proposed auths are ok. Ideally we would like to give the >>>>>>>>>> option to the user to register also with his email. >>>>>>>>>> >>>>>>>>>> The design you are mentioning looks good, you are in the right >>>>>>>>>> direction. >>>>>>>>>> >>>>>>>>>> The profile pic can be kept in the Sastix-CMS. It is a common >>>>>>>>>> approach we did also for our clients, when used this CMS. >>>>>>>>>> >>>>>>>>>> Grafana is ok also. Regarding your experiments with Kubernetes we >>>>>>>>>> propose MicroK8s, but it is up to you for your local tests. Ideally we want >>>>>>>>>> MicroK8s because we also contribute there, and it would be a good chance >>>>>>>>>> for you to be engaged in the future. >>>>>>>>>> >>>>>>>>>> SMS API is a luxury at this point, just integrate an email alerts >>>>>>>>>> service, and leave the SMS API as a last priority, if you have time. >>>>>>>>>> >>>>>>>>>> Regarding your potential open source idea, it would be very good >>>>>>>>>> to have. Keep an eye on what is currently out there and be careful not to >>>>>>>>>> get too overwhelmed and miss any GSOC deadlines/milestones. >>>>>>>>>> >>>>>>>>>> Best regards, >>>>>>>>>> Giannis >>>>>>>>>> -- >>>>>>>>>> Ioannis Skitsas >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Sun, Mar 28, 2021 at 2:37 PM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> >>>>>>>>>> wrote: >>>>>>>>>> >>>>>>>>>>> Hi Giannis >>>>>>>>>>> I did some research about how to implement frontend and setting >>>>>>>>>>> up monitoring using prometheus, have compiled all my findings into a doc >>>>>>>>>>> which I am attaching with this mail.Please let me know your views on it. >>>>>>>>>>> >>>>>>>>>>> Thanks & Regards >>>>>>>>>>> Sanskar Gupta >>>>>>>>>>> >>>>>>>>>>> On Sat, Mar 27, 2021 at 1:59 PM SANSKAR GUPTA <snskr95 [ at ] gmail [ dot ] com> >>>>>>>>>>> wrote: >>>>>>>>>>> >>>>>>>>>>>> Hi Giannis >>>>>>>>>>>> I have opened 2 issues in the cms repo, related to monitoring >>>>>>>>>>>> and frontend, we can discuss about specific topic in the issue itself , >>>>>>>>>>>> will be easy to track stuff. >>>>>>>>>>>> >>>>>>>>>>>> Thanks and regards >>>>>>>>>>>> Sanskar Gupta >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> On Fri, Mar 26, 2021 at 3:29 PM Giannis E. Skitsas < >>>>>>>>>>>> iskitsas [ at ] gmail [ dot ] com> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> Hi Sanskar, >>>>>>>>>>>>> >>>>>>>>>>>>> Do you really need to implement those endpoints for >>>>>>>>>>>>> monitoring? There are already solutions which we could integrate. Just take >>>>>>>>>>>>> into account that we want the whole solution to be dockerized and be able >>>>>>>>>>>>> to scale. We prefer to use Kubernetes, so ideally we want to use monitoring >>>>>>>>>>>>> tools that fit well there (Splunk, NetData, Prometheus, OpenMetrics, >>>>>>>>>>>>> DataDog etc). >>>>>>>>>>>>> >>>>>>>>>>>>> Regarding Kubernetes/Performance, yes we have used small >>>>>>>>>>>>> testbeds to see how autoscaler works, having videos been served from the >>>>>>>>>>>>> SastixCms-Kubernetes deployment. See this guide >>>>>>>>>>>>> <https://medium.com/@iskitsas/autoscale-a-java-cms-app-with-kubernetes-the-microk8s-approach-from-docker-to-kubernetes-7c021f7d8333> >>>>>>>>>>>>> for details. We have also used a dockerized approach for one of our >>>>>>>>>>>>> customers, with 3 large AWS instances serving content for ~5000 registered >>>>>>>>>>>>> users per daily basis. We have also parameterized the setup there a little >>>>>>>>>>>>> bit, to include NGINX proxies, for better caching and delivering of static >>>>>>>>>>>>> files. So, there are some tradeoffs you have to do when serving large >>>>>>>>>>>>> throughputs, and adopt your architecture accordingly. For this GSOC we are >>>>>>>>>>>>> ok to see something up to experiment level as a proof of concept. If you >>>>>>>>>>>>> are already familiar with those, we are open to discuss any improvements on >>>>>>>>>>>>> architecture level and support it with any new implementations. >>>>>>>>>>>>> >>>>>>>>>>>>> For any proposed enhancements, create github issues and we can >>>>>>>>>>>>> take the discussion from there. >>>>>>>>>>>>> >>>>>>>>>>>>> Thanks, >>>>>>>>>>>>> Giannis >>>>>>>>>>>>> -- >>>>>>>>>>>>> Ioannis Skitsas >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Fri, Mar 26, 2021 at 3:57 PM SANSKAR GUPTA < >>>>>>>>>>>>> snskr95 [ at ] gmail [ dot ] com> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Hi Giannis >>>>>>>>>>>>>> For GSOC stories , I was thinking if we can start by >>>>>>>>>>>>>> implementing necessary endpoints for disk usage , memory /cpu usage etc >>>>>>>>>>>>>> using websockets as it should be real time if I understand the >>>>>>>>>>>>>> functionality right. >>>>>>>>>>>>>> After that spring security can be integrated with creator and >>>>>>>>>>>>>> admin roles.We can discuss possibility of having more roles though >>>>>>>>>>>>>> I see you are using hazelcast for caching which according to >>>>>>>>>>>>>> me is a reasonable choice for distributed caching considering its a CMS. >>>>>>>>>>>>>> My doubts are: Has the performance of this app been tested on >>>>>>>>>>>>>> kubernetes, if yes what was the number of running instances? >>>>>>>>>>>>>> Also let me know if my plan to get started with backend >>>>>>>>>>>>>> enhancements sounds fine.We can then discuss the react app. >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> Thanks and regards >>>>>>>>>>>>>> Sanskar Gupta >>>>>>>>>>>>>> Masters Student at KTH Royal Institute of Technology, >>>>>>>>>>>>>> Stockholm >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> On Fri, Mar 26, 2021 at 9:41 AM Giannis E. Skitsas < >>>>>>>>>>>>>> iskitsas [ at ] gmail [ dot ] com> wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> Hi Sanskar, >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Thanks a lot for your feedback. We will review your PR. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Regarding the project see here >>>>>>>>>>>>>>> <https://ellak.gr/wiki/index.php?title=Google_Summer_of_Code_2021_proposed_ideas#Expected_Results> >>>>>>>>>>>>>>> for the expected results. We are open to any recommendations. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> GraphQL and the other "Features in pipeline" you see in the >>>>>>>>>>>>>>> README are not all required for GSOC. Specifically for GraphQL we wanted to >>>>>>>>>>>>>>> have this feature because it is easier for an integrator to understand our >>>>>>>>>>>>>>> APIs and data, and get exactly what you are asking for. You are free to >>>>>>>>>>>>>>> implement any of them if you feel you can deliver them in the proposed >>>>>>>>>>>>>>> timeline. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Let me know if you need any other information. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Best regards, >>>>>>>>>>>>>>> Giannis >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> -- >>>>>>>>>>>>>>> Ioannis Skitsas >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> On Thu, Mar 25, 2021 at 7:13 PM SANSKAR GUPTA < >>>>>>>>>>>>>>> snskr95 [ at ] gmail [ dot ] com> wrote: >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Hi Mentors >>>>>>>>>>>>>>>> This is regarding the cms project. >>>>>>>>>>>>>>>> First of all there was an issue in readme , the image url >>>>>>>>>>>>>>>> in sample request body to run the createResource endpoint was not working >>>>>>>>>>>>>>>> (was throwing 400) >>>>>>>>>>>>>>>> So I created an issue and a pull request for the same , >>>>>>>>>>>>>>>> requesting you to review it. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Pull Request Link <https://github.com/sastix/cms/pull/10> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Now regarding the project and possible enhancements about >>>>>>>>>>>>>>>> security, scalability and monitoring,I feel the following tech stack will >>>>>>>>>>>>>>>> be super beneficial and insanely fun to implement >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Frontend: Using material ui components to create a react >>>>>>>>>>>>>>>> dashboard for executing role based endpoints. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Security:Integrating spring security for resource access >>>>>>>>>>>>>>>> management >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Monitoring: Setting up a Grafana dashboard for constant >>>>>>>>>>>>>>>> monitoring API health and performance in TPS >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Scaling: Running the pods in a kubernetes pod with in place >>>>>>>>>>>>>>>> CI/CD pipeline , there will be no downtime during container lifecycle >>>>>>>>>>>>>>>> For CI/CD I would say I had some good experiences with >>>>>>>>>>>>>>>> CircleCI. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> For Archive API , any Relational DB should do the job where >>>>>>>>>>>>>>>> we can store meta data into separate columns or in a single string if >>>>>>>>>>>>>>>> redundancy is the deciding factor >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Not sure about the use case of GraphQL >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Request you to provide your thoughts on my choice of tech >>>>>>>>>>>>>>>> stack ,also request you to explain me the use case of graphQL in cms. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Thanks and Regards >>>>>>>>>>>>>>>> Sanskar Gupta >>>>>>>>>>>>>>>> Masters Student at KTH Royal Institute of Technology, >>>>>>>>>>>>>>>> Stockholm >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> ---- >>>>>>>>>>>>>>>> Λαμβάνετε αυτό το μήνυμα απο την λίστα: Λίστα αλληλογραφίας >>>>>>>>>>>>>>>> και συζητήσεων που απευθύνεται σε φοιτητές developers \& mentors έργων του >>>>>>>>>>>>>>>> Google Summer of Code - A discussion list for student developers and >>>>>>>>>>>>>>>> mentors of Google Summer of Code projects., >>>>>>>>>>>>>>>> https://lists.ellak.gr/gsoc-developers/listinfo.html >>>>>>>>>>>>>>>> Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό >>>>>>>>>>>>>>>> μήνυμα ηλ. ταχυδρομείου στη διεύθυνση < >>>>>>>>>>>>>>>> gsoc-developers+unsubscribe [ at ] ellak [ dot ] gr>. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>> -- >>>>> -- >>>>> Ioannis Skitsas >>>>> >>>>>
---- Λαμβάνετε αυτό το μήνυμα απο την λίστα: Λίστα αλληλογραφίας και συζητήσεων που απευθύνεται σε φοιτητές developers \& mentors έργων του Google Summer of Code - A discussion list for student developers and mentors of Google Summer of Code projects., https://lists.ellak.gr/gsoc-developers/listinfo.html Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό μήνυμα ηλ. ταχυδρομείου στη διεύθυνση <gsoc-developers+unsubscribe [ at ] ellak [ dot ] gr>.