[Okta Webinar] Learn how to a build a cloud-first strategyRegister Now

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 381
  • Last Modified:

Setup backup (or loadbalance) webserver in another city

Hi!

I have a web server setup in my city.  Even though I have an UPC, I am still concerned with webserver going offline.  So I am thinking about hosting a backup or loadbalancing server at a branch in another city.

Can anyone advise me on how to go about with this problem?

Greatly appretiated!

Danny
0
dangel
Asked:
dangel
  • 2
  • 2
  • 2
  • +1
1 Solution
 
lrmooreCommented:
Take a look at load-balancing devices such as Cisco's Localdirector for multiple servers at one location, or Distributed director for servers at multiple locations.

http://www.cisco.com/warp/public/cc/pd/cxsr/400/index.shtml
http://www.cisco.com/warp/public/cc/pd/cxsr/dd/index.shtml
0
 
t1n0m3nCommented:
Isn't Cisco's Local Director being discontinued?  (Specified as "End of Life")
I think that they are being replaced with "Content Service Switches"
http://www.cisco.com/warp/public/cc/general/qrg/eol_ai.htm
says that the last sale date is 3/1/98.
Here is a link to a content service switch.
http://www.cisco.com/warp/public/cc/pd/si/11000/

Interesting though the Distributed Director is not on the eol list.

The Local Director/Content switch is great for load balancing, on a local lan, but across the wan?
If the site that has the LD goes down, what does that give you?  Your service is still down even though you have the other link up.  I think that the solution would be to use DNS to give backup access to the link that is up.
http://www.f5.com/solutions/WAN/3DNSROI.html
http://www.akamai.com/en/resources/pdf/es_for_bc.pdf
http://www.autofailover.com/
0
 
t1n0m3nCommented:
Sounds to me like lrmoore's suggestion with the Distributed Director will work well with your situation. (and would be the one that I would pick)
The links above are other companies solutions as well.
Hope this helps.
0
Concerto's Cloud Advisory Services

Want to avoid the missteps to gaining all the benefits of the cloud? Learn more about the different assessment options from our Cloud Advisory team.

 
lrmooreCommented:
The local director 430 has been discontinued, but replaced by the 417 and 417G.

You are correct that if the site loses connectivity, it won't help. This is where distributed director comes in.

You could always resort to simple dns round-robin
http://www.webopedia.com/TERM/R/Round_Robin_DNS.html
0
 
samriCommented:
Another approach is to use front end accelerator server, and have this server to fetch pages from the two back-end server.

One approach is to use Apache as the front end, and use mod_rewrite to implement such.  There is a complete configuration example from Apache website to do that : http://httpd.apache.org/docs-2.0/misc/rewriteguide.html (scroll down to "Reverse Proxy" section).

This approach should work, but according to some, it might be a bit slow.  It's worth trying thought, and it's free (except the hardware and manpower to maintain it).

cheers.
0
 
scraig84Commented:
Samri - so what happens when the front-end server goes down?

I think the key point that people miss here is that if redundancy is truly important, you need to look at redundancy in many places.  If you add a load-balancer, you probably need 2 (1 in failover).  Also, there are redundant Internet links to look at requiring multiple routers (also a possibly more complex setup if you use multiple ISP's).  Also, you will probably need multiple layer 2 devices (hubs/switches) throughout this requiring multiple interfaces all around.  

So the question becomes - if you are really concerned with fault-tolerance, how far are you willing to go and how much are you willing to pay?

However, to answer your question about redundant servers, you may want to look at something cheaper than an extra device, such as Windows Load Balancing Service which is free (assuming you are on a Windows platform).
0
 
samriCommented:
scraig84,

Then the whole thing collapse :)  I guess something would end-up stop somewhere anyway.  Btw, you got a point there.

And it's true -- it all will depend on how *crucial* the uptime is, and how much $$ are to be invested, and how much $$$$ is to be gained.

and even if we had two (or more), the question would be.. what happened if all were down. :(

and again, my proposition is just an option, may not be the best (but sure a possible alternative).

cheers.
0

Featured Post

NFR key for Veeam Agent for Linux

Veeam is happy to provide a free NFR license for one year.  It allows for the non‑production use and valid for five workstations and two servers. Veeam Agent for Linux is a simple backup tool for your Linux installations, both on‑premises and in the public cloud.

  • 2
  • 2
  • 2
  • +1
Tackle projects and never again get stuck behind a technical roadblock.
Join Now