I am working on Highlevel multipart fileupload on aws sdk using java.But i need few clarification on below points .please help me!!
1 : What happens if there was network failure during uploading file parts?How do we re-upload ? will re-upload start from the beginning or will it get resume?
2 : Which provide better performance comparatively between low level and high level api
3 : any websites that provide Production ready code.
Thanks in advance..............
Resuming completly depends on how you have handled your implementation where you hold upload id and last part number that was uploaded. This is all you need to know to start resuming uploads in s3. That is the whle benfit of AWS S3 multipart Uploads.
Obviously High level API are good for your application if you really look forward to get progress of every upload you make. But that again depends on your application. for java sdk AWS Documentaion is too good.
Please go through AWS Java SDK Examples :https://docs.aws.amazon.com/AmazonS3/latest/dev/usingHLmpuJava.html
If you are implementaing in Java AWS Docs will be sufficient
Related
I'm looking for some solution design assistance, as an idea I thought I could mimic apache's httxt2dbm tool.
Use case:
We currently manage our 301 RewriteRule's in apache for our AEM environment. My co-worker has built a webform for end users to submit their redirect requests and this puts the form data in MongoDB.
Problem:
How can I convert JSON into a dbm hash file like the httxt2bdm tool does?
Any ideas for how to tackle this problem?
The end goal here is that users can fill in the webform, then a cronjob will run and update apache followed by restarting httpd.
I know about the map manager tool but it doesn't really fit the use case.
With the latest ForgeARKit-update-6-2018.1, I was trying to load my model in Unity, with the sample Unity scene 'loadAtStartup'. I can successfully load the sample models from 'Sandbox', but I couldn't load my model, which was uploaded through script 'test-2legged'.
Error message shows 504, it seems not reaching the service:
AsyncRequestCompleted The remote server returned an error: (504) Gateway Time-out.
UnityEngine.Debug:Log(Object)
Autodesk.Forge.ARKit.RequestQueueMgr:AsyncRequestCompleted(Object, AsyncCompletedEventArgs) (at Assets/Forge/CodeBase/RequestQueue.cs:322)
UnityEngine.UnitySynchronizationContext:ExecuteTasks()
Model URN:
dXJuOmFkc2sub2JqZWN0czpvcy5vYmplY3Q6bWFvbGlua3ppOHM3cnlvZWx4bjVndnR4bjcyZWc2N2l0dGp0a2MvMmZsb29yX0FyYy5pZmM=
[Update 23/4/2019]
I found that I can successfully load the same model with ForgeARKit-update-3-2017.1.2f1. I compare the Forge code in Unity. I think it has something to do with the service URL. The version 6 is fetching models from 'https://developer-api-beta.autodesk.io' while version 3 is fetching from 'https://developer-api.autodesk.io'. Meanwhile the shell script 'test-2legged' is uploading to the latter one ('https://developer-api.autodesk.io'). That is why it counldn'd find the resource. Question here is how can I upload model to the 'beta' ARKit? I tried modifying the URL in script 'test-2legged' but it doesn't work. Below screen-shot is the output of script 'test-2legged' when fetching from 'beta' ARkit. It seems model is uploaded successfully, but some parsing post-work failed. I guess the response format is also changed in the beta version. Is there a beta version of 'test-2legged' scripts (and other Scene Preparation scripts)?
Please comments, Thanks.
This is correct. My apologizes for this, I know we did not documented very well on the server changes.
This update6 assumes you are using the new server under beta right now. The scripts and update 3 are using the legacy server. Note that the 2 servers are not necessarily compatible and store the data in different places, so make sure to always use the same server in Unity as the one you used to prepare the scene. When we will switch everyone to the new server, we will transfer the data from the legacy server to the new server cloud storage.
The Update3 package will still be able to read scenes from the new server, as we make sure the old Unity code stays compatible.
Note as well you need to use SafeBase64 encoded string everywhere. I saw in your description that you are using base64 encoded (not safe). The new server will be more strict of parameters and format, so I encourage you to test your scripts/code on the beta server.
Last, I am working on a new Unity code update, and documentation which will be released next week. Make sure to use this version - it adds support for 3legged, automatic 2/3legged token refresh, and more. If you got scenes failing, please contact me directly and share your models and URN. I'll either test it on my development environment, or look into our log files for the reason to fail. My email address is my first name at autodesk.com
Thank you Cyrille for your help!! I am replying you here as it's easier to insert images.
I replaced the function 'xbase64encode()' with 'xbase64safeencode()', and now it works! However it seems for some model it still responds some error and in that case it cannot be loaded in Unity. (as the image below). I checked the script and I think all the encoding are using SafeBase64. Any clue of that? Or is that caused by my model?
BTW, the loading performance is greatly improved than the legacy version!! It looks almost the same as the web client. Huge Thanks for that!
Good to know that there is going to be an update next week. Yes I will test it and get back to you later.
I new to loop-back platform. I follow this tutorial loop-back official tutorial for third-party and this tutorial also a tutorial that i find.
What i want to do is to build a rest API using loop-back as one project and a client side using angular as another project then to wrap it for mobile. In this tutorials they are using loop-back passport for make the oath connection with Facebook.
now my questions are :
They save all the models involved in the authentication in the memory and not in the database. is it good approach? don't i need to save all the access-tokens, identities and credentials on the database ?
is there another tutorial that you can direct me to or you can explain me , where in loop-back i get the information return from Facebook after authentication because i want to save it in my user model ?
Thanks
1.) It's not recommended to save authentication-data in memory in production, you can easily change this in server/model-config.json (or model-config.production.json) with the variable "dataSource" for each model. So instead of "memory" change it to the name of your datasource.
2.) In the example you provided, the "customProfileToUser"-function seems like an excellent place to save eventual extra data to your own User-model.
I'm writing what is essentially a browser in Adobe AIR (ActionScript, not AJAX). A great bit of functionality to implement would be protocol handling. iTunes, for instance, handles itms protocols; when your friend sends you a link beginning with "itms://", it's going to launch iTunes as long as it's installed. Is there a way to write an AIR app (requiring AIR 2 would be fine) that can be the "handler" for a protocol in this way?
There is no way, programatically speaking, to specifically handle a particular protocol. However, there is InvokeEvent. InvokeEvent will be fired when the application is "invoked", either when it's explictly launched or if an associated file or URL is activated.
The process of associating your app with a particular file type or protocol scheme is separate and application-dependant. In iOS, for example, you would need to specify the protocol in Info.plist under CFBundleURLTypes/CFBundleURLSchemes.
Yes. You can use the URLLoader class to download data in binary form (URLLoader.BINARY) and then parse this as appropriate. See this CS3 documentation on working with external data.
http://www.patrick-heinzelmann.de/labs/lastfm/
I'm not sure exactly how it works and I don't see a way to download the app, so I can't even test it, but maybe it will help...
Check out this page. I am trying to find out the same thing, but I haven't found any solution to do it with just Air yet. Seems like you might need a custom installer to setup the correct registry entries, and a proxy application to "wash" the input to a correct format that then can start your application with the correct command line parameters. Hope this can be of any assistance.
Is it possible to multithread file uploads using http post requests (not ftp)? Are there any flash or java uploaders that take advantage of multithreaded uploads?
It is possible, and I'm sure there are several Flash and Java applets that supports it. It's just a matter of creating separate connections. Here's one of the results I found using Google:
http://www.uploadify.com/
HTML5 MultiFile uploads and this is example:
http://safron.su/playground/html5uploader/
P.S. This example is designed to load the picture.