Hi,
Can someone tell me how client can be connected to cloud? means is there any authentication kind of thing used by client and if it is then how it works?
And one more question is that is it possible to use same node controller for more than one cluster controllers or will there be any conflict as we need to pass multiple ssh key in that case to same nc?
Please reply ASAP . It's urgent...
Hello,
for the client authentication part, you can check the AWS documentation, since we are supporting the same mechanism.
Why would you want to put an NC on multiple cluster controller? Do you envision a specific reason to do so? And I'm not sure what you mean for ssh keys here: could you be a bit more specific?
cheers
graziano
Actually sir ,what i understood according to AWS authentication is , client also has to download credential from Web-UI. e.g. https://:8443/ what i thought is for this client need to know the username-password of Web-UI, which is not feasible as you should not allow client to change some configuration in it. Only cloud controller should know about it.
For single NC in multiple cluster:
Sir i just got that idea that whether it's possible to connect single NC with multiple Cluster or not. The other reason was that there is other group working on cloud and we both have to use same machine so obviously we'll setup NC in common m/c but i had doubt that it won't work. So can u clarify it ..
Sir i want to add one page on the CLC side , which contains machine id of all available NC and images stored.
and which allow client to select between the resources, so for that i have to do some coding so how to proceed? means how can i know what CLC does during registration of NC,CC and when request comes? which are the functions involved in this as according to my knowledge CLC,walrus,SC are made in JAVA and CC,NC in C. so how they interact. which are the system calls(and functions) used? In short code part during these process..
waiting for reply.
Hello,
hmm ... I'm not quite sure I still understand: part of the first time setup is to change the password for admin. If you are just a normal user,. you have to apply for the login, which means that you need to come up with a username password pair, then you need to be approved by the admin, and then you can only get your credentials, not modify the configuration. Why do you think that normal user can change configuration?
NC on multiple cluster is a problem: you need to have credentials setup correctly to have the CC talking to the NC, and those credentials are generated on the fly when the cluster is added. I assume you can try to force the same credentials everywhere, but this will for sure cause confusion in your setup.
At the moment we don't have any good documentation on the code: we just have high level explanation (which we posted). As of now the only documentation is the code itself :)
cheers
graziano
For the first part what i though that client must know the username/password of web-ui to download credentials and that username/password are same as used by clc so it's not good practice but as u said client username/password must be approved by clc so it's different from one used by clc.So that doubt is cleared.
For the last part,i like to give an example to make it clear.
Suppose i have three different versions of mysql(or any other service that i want to provide) on clc now when client requests , client requests only with mysql without specifying version. So clc must then provide version info to client and when client selects anyone of them then that selected version must be run by clc on particular NC. So for that i need to do some coding between client and clc so how to proceed?
Another thing is along with version selection i also want the client to select the NC to run service so for that i need clc-cc(or nc) coding. So how to proceed in that also?
Hello,
Eucalyptus will allow to run instances: what you put into the instance is up to you and whatever you want to do with it, is again, up to you (or the appoication). In your case, if you want to provide multiple version of one service, you will have to understand how to do that: it's not different than having just bought say 3 new servers, and now, after you install the OS on it. you are free to do what you want with them.
Since it looks like you are new, we have a User Guide in the documentation and some quick first steps in the community wiki you may want to read. Remember that all data in the instance will disappear when you terminate it: if you want persistent data you may want to look into either EBS or walrus (S3).
cheers
graziano