Solved

Join a server to the network from amazon EC2

Posted on 2016-09-02
19
108 Views
Last Modified: 2016-09-11
I have a window server on amazon EC2 and want to join it to the network in the office so it can replace the server in the office and provide the all services a server provides to all workstations and other devices. Domain, DNS, DHCP, Shared folders, GOP.....

Which service from amazon would do this job fast and ez, (only need it for temporarily)

Please advice
0
Comment
Question by:Abraham Deutsch
  • 7
  • 6
  • 3
  • +2
19 Comments
 
LVL 36

Assisted Solution

by:ArneLovius
ArneLovius earned 125 total points
ID: 41782584
It would be useful if you could describe why you want to do this, there may be other alternatives that could provide your end result in a different way.

That said...

There is no single service from Amazon to do this, it would not be very difficult, but I would not class it as easy.

First of all, you would need a private connection to your EC2 instance(s), this is simplest with a site to site VPN between your EC2 instance(s) on Amazon and the network in the office. You could use the "AWS hardware VPN" or "Software VPN" (which woudl mean you running a VPN "appliance" on another EC2 instance) options http://docs.aws.amazon.com/AmazonVPC/latest/UserGuide/vpn-connections.html

Once you have a site to site VPN in place, you can run promote an EC2 instance to a Domain Controller for replication AD and DNS etc.

For DHCP, you then would need something in the office with a DHCP helper to forward DHCP requests across the VPN to the server on EC2

If you are currently using file shares that reference the server name, I would suggest migrating to a domain based DFS namespace https://en.wikipedia.org/wiki/Distributed_File_System_(Microsoft) https://technet.microsoft.com/en-us/library/cc753479(v=ws.10).aspx and using DFS replication to "copy" all file share content. Alternatively, you could use tools such as robocopy or rsync to copy data to your EC2 instance and manually change the share names on workstations.
1
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41783808
I am looking for a backup solution to offer my clients that will have the minimum down time in case of a disaster.

What I came up, do a daily backup to AWS when a disaster happens get the server from AWS running in EC2 and with the service (which in question) get the server into the network so the office is up and running.
0
 
LVL 36

Expert Comment

by:ArneLovius
ID: 41784491
You have a nice high level idea, but the devil is in the details, being able to spin up an offsite replica of an onsite server is quite a complex problem, this would usually be done bay taking an image (can be VSS snapshot) based backup to an "offline" disk offsite.

Take a look at Datto as an example.

If all of your clients used hyper-v for running their servers, then you could look at hyper-v replication over the internet using certificates for auth.

Bear in mind, that to be able to keep an offsite server replica up to date, requires bandwidth roughly equivalent to the amount of data change on the source...
0
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41784513
I am currently looking on this product http://www.cloudberrylab.com/ but I told them that the main feature is missing bringing it in to the network. And thy claim it's something Amazon offers out of the box. Which is not the way I see it and seems you as well.
0
 
LVL 36

Expert Comment

by:ArneLovius
ID: 41784546
Cloudberry backup can restore to EC2 (or baremetal or hyper-v or vmware) , and the backups would continue, you could then restore again to bare metal (or hyper-v or vmare), but there is no way to "move" the EC2 instance.

In any event, you would still need the VPN to AWS as I detailed previously.

As an alternative idea, You could keep a suitably specified hyper-v/vmware host ready to have a backup restored to which you could then take onsite.

Have you calculated the AWS S3 costs of using cloudberry for one of your clients ? It may be significantly more than you expect...
0
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41784755
With the alternative way I will only able to offer onside backup as it would be impossible from my side to handle this much of bandwidth (once the clientele base starts growing)

Re cost amazing AWS is only 0.03 a GB and cloud Barry is a one time fee of $59 par year but the problem is when there is a need for a restore if it I'll do it to my machine the download will take a significant amount of time, and can take easy 24 hours to be back in production.
0
 
LVL 36

Expert Comment

by:ArneLovius
ID: 41784902
Do remember that the $0.03 per month is for data stored on S3, so as an exampe, if you had a very small file server of say 100GB, your base cost would be only $3 per month if no files change, if you have 10Gb of changed files per day, and the smallest billing period for Amazon is one month, so your additional cost will be $0.30 * 20 working days in a month = $6, so your actual cost would be $9 per month, three times higher that just the base storage. if you had 1TB on the file server with the same percentage of file change, then your costs go from $30 per month to $90 per month, or $1080 per year.

If you compare to using windows backup with external USB disks, the pricing may start to lose some of its initial appeal.
0
 

Expert Comment

by:Anton Zorin
ID: 41786041
Why don't to use Cloudberry for onsite backups? In case of disaster you can restore a server locally (in other words fast enough) and if the entire site is burned/flooded/(insert your option) you can restore it to EC2 and proceed with functioning.
Having AD/DNS/DHCP/Fileshare server on EC2 VM will require high I/O and thus expensive instance with a drive (for file sharing) big enough to store all the data but it will be EBS and thus $0.09/GB/mo.
You have to get all the numbers down in order to see the full picture.
0
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41786109
I agree but in order in case of a total onsite disaster to be able to do a restore to EC2 it needs to be first backed up to AWS [without a EC2 running and only use EC2 if needed] correct?

but again comes up the question how to join the EC2 to the network.

but if I understand you correctly a local backup should be enough for a server failure and even with a NAS drive [not sure yet on the model which sport] have the server up in running as a VM. And in case  total onsite disaster to recover relocate and set up all other stuff will give enough time to download from AWS to do a local restore even with large amount of date. Correct?

PS wondering what is the point of them putting in so much effort in restore to EC2 when i don't see how to make use of it.
0
Netscaler Common Configuration How To guides

If you use NetScaler you will want to see these guides. The NetScaler How To Guides show administrators how to get NetScaler up and configured by providing instructions for common scenarios and some not so common ones.

 
LVL 22

Assisted Solution

by:David Atkin
David Atkin earned 125 total points
ID: 41786337
Just to throw another vendor out there.

Datto have a system that does pretty much what you're after I think.   It backups up to a NAS as a VM.  In the event of a server failure you can start the VM on the NAS.  In addition, you can replicate the backups off site.  In a total site failure situation you can start the VM in the Datto cloud.
0
 
LVL 36

Expert Comment

by:ArneLovius
ID: 41786614
You wouldn't be joining the restored instance on EC2 to the domain, in a single server environment it would _be_ the domain...

Once restored, you would have to change its IP address, and have a method of providing a VPN from the remote site to access it. DHCP could be handled by the router/firewall.
0
 

Expert Comment

by:Anton Zorin
ID: 41787596
As far as I understand you can't have a server in EC2 with no DHCP configured thus addressing is not an issue. Computer names may cause some issue though.
0
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41787636
As well in order for the workstation to access their files on the server in must be part of the network.

So assuming that in order to have the office up and running the server in EC2 will need to be joined to the network, I come back to initial question what service from Amazon is needed and is it something out of the box as cloud Barry claims so it's fair to promise a client that in case of a disaster his office will be operational in a short period of time?
0
 

Expert Comment

by:Anton Zorin
ID: 41787713
In case of workstations you can use Amazon Workspaces and just RDP to them from your desktops.
I think you can promise only what's achievable is both RTO and RPO described properly.
0
 
LVL 1

Author Comment

by:Abraham Deutsch
ID: 41787728
Sorry I did not explain my self correctly.

What I was referring to is. In case the server fails its only the server all workstations in the office are in perfect condition re hardware but in order to be in production in need the server for to have access to its file on the server. in order this to happen the server need to be on the network.

PS I has set a appointment with Datto  to discuss their product.
0
 
LVL 14

Accepted Solution

by:
Kaffiend earned 250 total points
ID: 41788728
AWS or Azure, or whatever-you-choose, *might* work if you had enough bandwidth in the office.  If this office doesn't have very good connectivity, it would probably be a very painful experience for the end-users.

Not exactly what you asked for, but if all you need is a quick way to recover from a crashed server:
- virtualize the existing one
- use virtual-aware backup software (most decent ones will provide "whole server" recovery) and periodically transfer copies of these backups to somewhere off-site
- in case of server crash (and the rest of the office is in good shape) you can be up and running within minutes of you getting there - a decent spare PC, with a big SSD would work in a pinch - it would probably be faster than the existing server if this server is in danger of hardware failure)
0
 
LVL 36

Expert Comment

by:ArneLovius
ID: 41789184
as I said in my first answer, you would need a VPN to connect the instance of the server running on AWS to the site...
0
 
LVL 14

Expert Comment

by:Kaffiend
ID: 41789887
+1 to ArneLovius' comment above

Too much work to set up in the event you actually needed it.  And it's going to be a pressure situation, with everyone looking over your shoulder as you set it all up - VPN, your hardware firewall, DNS, DHCP, AD, etc etc
0
 
LVL 1

Author Closing Comment

by:Abraham Deutsch
ID: 41793373
although ArneLovius answered my initial question, but the "solution" to accomplish my goal was bast answered by Kaffiend and assisted solution by David Atkin.

Thank you all
0

Featured Post

Enabling OSINT in Activity Based Intelligence

Activity based intelligence (ABI) requires access to all available sources of data. Recorded Future allows analysts to observe structured data on the open, deep, and dark web.

Join & Write a Comment

When the confidentiality and security of your data is a must, trust the highly encrypted cloud fax portfolio used by 12 million businesses worldwide, including nearly half of the Fortune 500.
Data center, now-a-days, is referred as the home of all the advanced technologies. In-fact, most of the businesses are now establishing their entire organizational structure around the IT capabilities.
This Micro Tutorial will explain how to export DynamoDB tables in Amazon Web Services.
After creating this article (http://www.experts-exchange.com/articles/23699/Setup-Mikrotik-routers-with-OSPF.html), I decided to make a video (no audio) to show you how to configure the routers and run some trace routes and pings between the 7 sites…

708 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

15 Experts available now in Live!

Get 1:1 Help Now