I'm testing Google Cloud DNS.
I have 3 A NAMES in my test zone that all resolve to the same DNS name.
The servers are in Australia, Northwest US, and India.
I am located in Northwest US.
When I do a tracert on my DNS name, it goes to the IP in the Australian data center.
From me to the Northwest US data center has much lower latency.
Why does it pick Australia? I thought latency based routing was "built-in" to an anycast network.
thanks in advance
Google Cloud DNS does not do latency based routing.
The Cloud DNS nameservers are on anycast network, so your DNS queries will be routed to the "nearest" nameserver. See https://cloud.google.com/dns/what-is-cloud-dns#performance_and_timing. Your servers' ip addresses, however, are not on that network.
To my knowledge Google Cloud DNS doesn't do lowest latency routing. AWS has regions in the places you mentioned, so it might be worth checking out Route 53. I primarily use GCE for compute resources, but generally use Route 53 for DNS. http://aws.amazon.com/blogs/aws/latency-based-multi-region-routing-now-available-for-aws/
http://aws.amazon.com/route53/faqs/#Route_53_anycast
If you have 3 A records on the same domain. The nameserver will ALWAYS return all 3 IPs.
Then your computer just pick one of them to ping/traceroute. Same for browsers(browsers will try another if the current one is down).
If you want the nameserver return the nearest server, use AWS Route 53's latency/geo based function. Google Cloud DNS don't have such function so far.
However if you have all your servers at Google, you may want to try their Load Balancer + CDN. The overall latency and performance is amazing, just like the speed your credit card is burning. lol
Related
For our SaaS app, we're allowing customers to point their domain name to our server.
The plan right now is to simply hand out one of our AWS elastic IP addresses for them to point their domain to. The elastic IP address would essentially be pointed to a EC2 instance web-server...and maybe a load balancer in time (if traffic demands it!).
The user would specify what their domain is in our app, and we'd be able to resolve the host name coming in as their app.
My concern is the longevity of this solution. This IP cannot change. And we'll certainly be tied to AWS if we go this route.
(Note: Being a 1-2 person startup, standing up a data-center is more than likely no-go, and we hope to use AWS or Azure).
What solutions would make this IP address -> SaaS Web Server concept last in the long run, with flexibility, and as minor of a tie as possible to a cloud provider?
With running the risk of asking "what is the best way to do this"...what's the best way to do this, keeping in mind longevity and small opt-in to a cloud provider?
You can't point an IP address to a load balancer, so this seems like a very bad idea. You need your own domain/subdomain that clients can point their domains/subdomains to via a CNAME record on their end. Then if the location of your service ever changes you just have to update your domain record and their DNS records will continue to be correct.
This issue came up as a result of SEO concerns, but having done some further research, it seems Google feels that IP/hosting location are now a weak signal for ranking, at best. So now I'm just curious, as I'm only familiar with networking on a basic level.
I have several sites hosted in the europe-west-1 region. Each site is on a compute engine instance with an external static IP assigned. I can ping the domains/IPs and then have my colleague in the UK ping the same and based on response time it's clear that the IP is ultimately resolving in Europe (probably Dublin, Ireland where it should be). But a DNS lookup of the same domain/IP lists the IP in Mountain View, CA? It always comes out like this: xxx.xxx.xxx.xxx.bc.googleusercontent.com. Is this Google acting like an ISP, and then the routing to Europe is behind the scenes? Why do none of the IPs show as resolving in the data center where the instances are hosted?
It sounds like you are conflating three concepts:
The domain registration for the googleusercontent.com reverse DNS
The SWIP record for the subnet
The routing decision to reach an IP
All three are independent. All GCE instance IP addresses have a reverse DNS entry that maps to xxx.xxx.xxx.xxx.bc.googleusercontent.com. This domain is under Google's control and therefore registered to the HQ in Mountain View.
The SWIP record / WHOIS entry denotes the administrative ownership of an IP address resp. its subnet. It's therefore also registered to the HQ in Mountain View.
Both of these do not reflect anything about physical location of the machine answering packets to an IP address nor the decisions on how packets are routed to the destination.
Google has a global network. Packets to a GCE instance will cross over to Google's network relatively close to the client. Since Google maintains a lot of peerings with ISPs worldwide, most of the time your packets will end up on Google's network directly from your ISP.
If you run a traceroute to your instance, you might see hops with airport codes in their reverse DNS names, especially when traversing peering points. The hops internal to Google usually do not hint any further at geographical location.
And finally, when the "proximity" or location of an IP address is discussed, most of the time the relevant metric is the latency or network distance to the host - not the geographical distance. (Although geographical distance sets a lower bound for latency as packets cannot go faster than the speed of light)
I have a webapplication in multiple Regions in the Azure Cloud and i'm using the Traffic Manager in Performance mode zu redirect the user to the closest Region.
What's concerning me is the following:
With this site https://www.whatsmydns.net i checked my Webapplication to see, which Datacenter is selected.
The funny thing is, that people from California gets redirected to the server in Westeurope but there is a Server in US Central too.
So from the site of the traffic manager the ping to the europe server is faster then to US central.
But i believe, that the difference between these too can not be high...
Now i have the fear, that it can happen that a user jumps between US Central and Europe all the time because he is in such a zone where the latencies to the available servers are nearly identical.
I also store files in a Azure Storage account in each region. If the user now jumps, i would have to transfer these files between the regions all the time...
So i was wondering if there is a possibility to redirect the user by his GEOIp to a specific region than by latency?
One of the benefit of the traffic manager is in my eyes that i can use one domain for all regions...
the only solution for my problem i can think of is a own cloudservice which replaces the traffic manager and redirects the user to the different regions by their IP like us-center.DOMAIN.com, we-eu.DOMAIN.com etc...
Are there any other solutions?
Thanks for your help!
Br,
metabolic
If you believe Traffic Manager is routing queries incorrectly, that should be raised with Azure Support.
Traffic Manager 'Performance' mode routing is based on an internal 'IP address to Azure data center latency map. The source IP of the DNS query (which is typically the IP of your DNS server) is looked up in the map to determine which Azure location will offer the best performance. There is an implicit assumption that the IP address of the DNS server is a good proxy for the location of the end user.
The 'Performance' mode in Azure Traffic Manager is deterministic. Identical queries from the same address will be routed consistently. The only exception is that routing may change during occasional map updates, which affect only a small %age of the IP address space.
A more common cause of routing changes is customers moving from place to place. For example, during travel, or simply by picking up a Wifi network that uses a DNS service in a different location, with a different IP address.
A Geo-IP based routing is not currently supported by Traffic Manager. However, please note that it would work in the same way as the 'performance' routing, just that it would use a different map. Users could still be routed to different locations as a result of map updates or changing DNS servers.
As you describe, if your application requires a strong, un-violable association between a user and a region, one option is re-direct users at the application level (e.g. via HTTP 302).
We have 2 servers hosting a particular service on google cloud. How to do a simple round-robin DNS configuration to distribute the load?
According to this thread Google Cloud DNS does not support round-robin.
You can set up DNS round robin with Cloud DNS simply by adding more than one IP address to your DNS record.
You might want to look into Google Compute Engine's Load Balancing options. This will allow you to have one IP address that sends traffic to your two servers. This has a few advantages, including that you can configure it to automatically stop sending traffic to an instance if it fails a health check.
Fast question is it possible to have Azure Traffic Manager
I would like to rent dedicated servers in 3rd party suppler and to load balancer from Azure
Question 1:
Can I setup this scenario? and use the load balancer from Azure?
Question 2:
Will I pay Outgoing bandwidth
Question 3:
Will you share for website with 10 000 000 page views per month how much you pay for DNS look ups as average.
Question 4 please suggest same service competitors... Google, Amazon, Rackspace I already know
The link you provided to the article already answers #1 and #3. Yes you can set this up. Billing is done by DNS lookup at $0.75 per million lookup, so your 10m page views will cost at most $7.50, but this isn't taking into consideration DNS caching which will drastically lower this (already very low) cost.
Question 2 is not an Azure Traffic Manager related question. No bandwidth goes through ATM so there is no charge. I am sure you will pay bandwidth charges with whatever 3rd party datacenter provider you are going to use.
I don't understand question 4. What do you want suggestions for? A cloud provider? There are lots of good ones but it depends on your scenario.
Azure Traffic Manager is a DNS routing system. It is similar to the routing features of AWS Route 53 (although Route 53 is a more full-featured DNS system).
Azure Traffic Manager uses DNS to point incoming traffic to different endpoints, which can be either within Azure or external urls. Because it uses DNS, it doesn't actually see any of the data itself, it just translates something like myapp.trafficmanager.net to 'webserver1.example.comorwebserver2.example.com` based on your rules and setup.
You can use round-robin, weighted or performance (which directs to the geographically closest address you have setup). You can further use Azure's DNS or another DNS system to use your own (sub)domain to CNAME to the trafficmanager.net domain name.
Load balancers like Azure Load Balancers and Amazon's Elastic Load Balancers are used to actually spread the traffic itself to different machines or services. Each work only with services hosted with the cloud provider so Azure Load Balancers can be used to load balance Azure VM's but not some servers you have hosted elsewhere.
Load balancers have bandwidth charges because they actually pass through the traffic. Azure Traffic Manager just has DNS query charges because that's all it does.
In your case, yes you can use Azure Traffic Manager to point to several external endpoints for your dedicated servers. You can also nest Traffic Manager profiles so that you can first use geo-location then round-robin. Azure Traffic Manager does support basic http/https monitoring to make sure the endpoint is still active.
Because it is based on DNS, there will always be a lag between changes with the TTL value and how clients cache DNS addresses. This is inherent with all DNS routing. To be extra safe, you can use Azure Traffic Manager to route to your datacenter and then run your own load balancing software locally to spread the load among servers.