Whether OpenEBS can help me to implement content delivery service (only for audio files) to replace S3?
This is a very broad question.
Short answer: Yes. OpenEBS is a way to for you to use Volumes in Kubernetes. To implement a Content Delivery Service you need some kind of application with a storage backend. OpenEBS allows you to be that storage backend.
Related
Reading several posts on adding images to QnAMaker KBs, these URLs all have to be externally accessible, like on Google images. How can I add an image available on my company intranet?
I know this isn't a direct answer to your question, but the easiest way to do this may be to copy and host the images on an Azure App Service. It creates more overhead, especially if these images get updated with any frequency.
I'm assuming you already have an App Service Plan running your QnA Maker Bot. You can create a new App Service on the same plan so that you don't incur any incremental cost. Just upload the images onto the app service and you'll be able to access them from QnA Maker or any other application. There are a lot of ways to get the images there. Personally I use Azure DevOps Repos to manage the code/images, but any source control app could work; I'd recommend using something Azure Deployment Center can integrate with directly. You can use the Kudu build service to deploy the web app, no Azure Release Pipeline or other deployment methods necessary.
I'm wondering though, if your users are internal to the company, couldn't you link to the images on your intranet anyway? If they are behind your firewall the images should still load even if Azure itself doesn't have access to it (because all you're providing in QnA Maker is the URL). I'm assuming you're asking because the bot is for users external to your company, but I figured I'd throw that out there.
I need to do cloud based recording. I need to take all video+audio streams of all clients. Then push to AWS S3 or Gcloud bucket or Agora storage itself. I need to do it through API since I have some criteria and also I need to organize the uploaded data in specific folder names.
I did not see any specific function in agora service to do so. Should I take copy of localstream and handle it on my own? If so, I hope it does not come in the way of agora functionality.. Please share if any sample project. Just to emphasize- I am not looking for on-premise SDK. I need to do it through REST. I use angular 8. I could write my own REST server (to deal cloud of my choice) if really needed.
Background study done-
In agora.IO documentation, I see lot of documentation about on-premise recording. I did not see any reference for cloud based recording. I checked the agora.IO documentation and also angular code which includes agora service component. Angular sample works fine. But I did not understand how to take the stream and upload to my cloud storage. I am concerned that it might impact the streaming and playback of agora. To handle this , I thought you might have specific guideline or library.
Thanks
There is a new Cloud Recording API in beta that Agora is offering that can connect to the channel and record the streams (configured by RESTful API) directly to your S3 bucket, currently GCP is not supported.
This is an add-on feature so it's not enabled by default and not available to be enabled through the dashboard, it needs to be enabled by the Agora team. Please Join the Agora Community Slack and send me a direct message with the AppID for which you wish to enable with this feature and I can get you into the beta program.
I am new with Software defined storage. I was working on possible ways of integration of different object storage implementations like AWS S3, Openstack Swift with Ceph.
I am wondering if I can use ceph API calls to store objects in Openstack swift?
The document here specifies that I can use Swift APIs to store objects in ceph (OSDs), but is it possible the other way around?
Thanks for any help in advance.
Exaclty. Ceph Rados Gateway is a full replacement of Swift Object storage with some missing functionality as described here.
http://docs.ceph.com/docs/master/radosgw/swift/
To make it clear, in order to store objects in Ceph Cluster you should just setup Rados Gateway and that's it. You will have two options to operate either using Swift API or S3 API.
Rados Gateway also supports S3 API to store objects. You can even configure RGW to access containers like S3 bucket subdomains. Ex: #bucketname#.s3.example.com
After going trough this link on API versioning, i am trying to adapt the version through the headers and deploy it of AWS, but it looks like AWS does not support any custom verndor specific MIME types(link here)
Any idea how to achieve the versioning in this scenario where you do not have control over the MIME type...or any insights on how to do API versioning when using AWS?
My suggestion for cloud based service is "v1.api.mysite.com". It provides good flexibility. More details and reasoning here - http://timanovsky.wordpress.com/2013/01/28/api-versioning-scheme-for-cloud-rest-services/
I have gone through the documentation of Amazon mechanical Turk. I find that there are options to create/monitor and expire the HITs. All the API's are the requester API's. Are there any worker API's available to select a task or indicate the completion of task?
All of the APIs are currently for use on the requester side of mTurk. Workers must interact with mTurk via the web interface.
I'm not sure if they are trying to protect their system/users/IP, or if this is just an instance of mTurk starting to show its age. It's one of the oldest AWS services and launched all the way back in 2005!