Failover Plan - If One Server Is Down, Pick The Second One
Apr 23, 2009
So, there is my example. I have two servers in different locations. First one is the main server with all the data and the second one has just simple notice like "We will be back soon".
Now, if I set the primary DNS server on main server and secondary DNS on second one (but with different A records), will users actually get the server two if main server is down?
Will this work or not? I want 100% access to the site, at least to the server with only notice (I don't need data from main server). I can set the round robin, but I don't need load balancing (actually I want access only main server), but just if main server is down to go on the second one.
I like to create some service plans using the cli-tools, /usr/local/psa/bin/service_plan.I am able to create a service plan, but I'm unable to create a service plan inside a reseller plan. For example I cannot "tell" the service_plan script to add the created serviceplan to a reseller plan. Is it possible to create a serviceplan inside a reseller plan, using the cli?
I host my own server but need to get a failover server in the case it goes down. Does anyone know of any companies that host failover servers that will run off my own server?
I'd like to create a failover server, a complete and identical copy (as much as possible) of my primary server, which becomes "primary" when the first server dies.
I've been reading a lot about this subject and i've seen an already working method which uses rsync and MySQL replication.
Has anyone managed to create such a system? is there someone with experience or can point me to a relevant web site?
Now I want to make a status subdomain on another server.
ibeat.org on one - status.ibeat.org on another.
I've edited the master zone on server one and added status.ibeat.org. and www.status.ibeat.org. ... and pointed it with A and CNAME to the other server and related IP...
I have also created the subdomain on the second server.
So my question is what should I do to get this working? This isn't!
I want to have a status link on the ibeat site ... and also:
... if server one with ibeat.org fails make it open status.ibeat.org?
Should be pretty simple - but can't get my head wrapped around it.
I'm going to be ordering a new dedicated Linux server w/ cPanel/WHM and was wondering how one goes about determining which OS is best suited for their needs? The options I am given are below:
OS: CentOS Enterprise Linux, Version 4 OS: CentOS Enterprise Linux - x64, Version 4 OS: Centos Enterprise Linux, Version 5 OS: Centos Enterprise Linux - x86_64, Version 5 OS: Red Hat Enterprise Linux, Version 4 OS: Redhat Enterprise Linux - x64, Version 4 OS: Red Hat Enterprise Linux, Version 5 OS: Red Hat Enterprise Linux - x86_64, Version 5
This is a tough question cause the distro you choose makes all the difference.
So my question now is if your choices were:
Centos 4 or 5 (32bit or 64bit) Ubuntu (32bit) Debian (32bit)
which would you pick and why?
on a side note I believe that the distro that gets picked the most is the one that has the most servers and if someone were to pick one thats not favored by many then he/she may get a faster server cause there wont be so many VPSs on it. Correct in this assumption?
I am trying to do a fail over solution with round robin dns. Our dns is served by windows and our web servers are setup with Linux..
I know round-robin does not by default do a fail over, however my understanding is that a script can be used to remove the failed server for dns, is anyone aware of something that will do this for windows?
I noticed with every new cent os server there are different mirrors for cent os repos.Since on each cent os repo file is same,how does it pick which mirror it pick?I ask beacuse latest server has picked one extremly slow edu mirror which is not just slow,it also timeout,and it may extend update to 2 hours instead 5 minutes which usually will take.So how do i change base mirror on base cent os repo?
I've used two different shared hosts in the past. First host had little to no support and got slower and slower as time went on. Second host, has been fine for my personal blog and portfolio site, but the load time is way slow at certain times of the day.
How in the world do you find a shared host that is actually going to be fast? I know you don't want a host that is over-selling, but how do you know a host is overselling? I assume any host that is less than $10/mo and has unlimited everything is going to be oversold, but how do you find one that is going to be responsive. Trial and Error? I've been trying to read the reviews on sitepoint, but I feel like a lot of the time the reviewer is unaware of site load times.
I don't mind doing the research myself, but I can't figure out where to look or how to look.
All I want from my host is fast load times, all the time. I don't have huge traffic, I do have a lot of files, and it's very important when someone comes to my site it loads as fast as possible.
Obviously cheaper is nicer, but I don't mind spending a little extra to have a server that loads fast.
I run a site with about 1,000,000 unique visitors per month and recents server failures made me decide to get a failover server to minimize downtime. My goal wasn't to get 99.999% uptime but to be able to be back on track after a failure in a "reasonable" amount of time. After evaluating several solutions, I decided to go with DNS failover. Here's how the setup work:
1) mydomain.com points to main server with a very low TTL (time to live) 2) failover server replicates data from main server 3) when main server goes down, mydomain.com is changed to point to failover server
The drawback is the DNS propagation time since some DNS servers don't honor the TTL and there is some caching happening on the user's machine and browser. I looked for empirical data to gauge the extent of the problem but couldn't find any so I decided to setup my own experiment.
The Experiment ==============
I start with mydomain.com pointing to the main server with a TTL of 1800 seconds (1/2 hour). I then change it to point to the failover server which simply port forwards to the main server. On the main server, I periodically compute the percentage of requests coming from the failover server which gives me the percentage of people for which the DNS change has propagated.
I made the DNS change at exactly 16:04 on 06/21/06 and here are the percentage of propagated users:
So even after 18 hours, there is still a certain percentage of users going to the old server so DNS failover is obviously not a 99.999% uptime solution. However, since more than 90% of the users are propagated in the first hour, the solution works well enough for me.
Im setting up a website that will facilitate people to be able to upload and download files.
As you can imagine a lot of people will be downloading so i need unlimited bandwith.
Also require maybe 1tb of diskspace with the option to expand wen needed.
Im not sure how cpu intensive this is but if people can recommend me specs, companies i should be looking for with links that would be very much appreciated.
we are looking for a suggestion on a best practice to help a customer recover from being hacked. One thought is that the client should be put on a new server and allowed to rebuild from there. The concern is that there are so many web app breaches that the server often gets rehacked because the bad code is still in the web app.
There is so much information on disaster recovery and backing up one's server, that I'm getting glassy-eyed trying to take it in. Maybe if I became an actual case study, and get some "group think" help, this thread could benefit many others in a similar situation.
Current Situation:
1. I'm a small hosting company, 5 years in existence, with about 350 clients. www.mlhi.net
2. Dedicated Linux server, PLESK CP w/unlimited domains license, fully managed at HostNexus (great guys). It does not have a RAID array (used to have that at Rackspace) but it does have a backup drive that everything is backed up to with a cron job every night.
3) In addition I have a Linux Sys Admin on retainer, www.linuxbox.co.uk (he is better than excellent). Two years of excellent server maintenance and security on top of the managed service I get at HostNexus.
4) I just bought a VPS plan at JauguarPC.com after much research (a lot of it here at WHT) and as they say "so far so good" with the ease of dealing with them. I have not setup anything there yet- just got the VPS provisioned a few days ago.
Fears and Concerns:
1. Data center destroyed/ my server burns up (including backup drive) etc etc.
2. DDOS attack (which did hit this data center a few months ago and I was down for hours)
3. If I had to FTP everything back to another server from my local, at 18 GB, it's not too cool.
Want to do this:
1. I want my Sys Admin to run a backup copy (and incrementals every night) to an identically configured VPS server at JaguarPC. Both servers are now running identical PLESK 8.4.
2. I want the fastest recovery possible without spending a ton of money. I know this means I don't get an "instant" recovery, but recovery within 24 hours is more than OK. None of my customers are ecommerce... just brochureware sites.
My "I'm not an expert" plan:
1. If primary server goes bye-bye forever, I can login to my BulkRegister/Enom account and change the child nameserver IPs to the IP's of the VPS. In 24 hours or less, every request for the nameservers would then be routed to the new server.
2. I can create an A record on every domain like www2.johndoeinsurance.com that would point to the IP at the VPS, so I can ease my mind anytime I want to make sure everything is safe and sound on the second server, and ready to go in an emergency.
How do I configure the DNS?
I control dns at Enom for about two-thirds of my customers. I have ALL domains pointed to ns.mlhi.net and ns2.mlhi.net. Here are my options??
1. I create two more child nameservers... ns3 and ns4 and have then pointed to the IPs at the new server, then update all the domains I control. The rest of the customers I can email and ask them to add the additional nameservers. I know... good luck on them doing it.
2. I change the ns2 IP to go to the new server. And I make sure when I make edits on a website during the day that I FTP to both servers.
3. I don't have any nameservers assigned to the new server. I just change the IP on the existing nameservers in the event of an emergency.
I am doing migration of plesk 11 from one server to another server on both server version is same but after migration I cant find any service plan which is there on old server.how can i copy / migrate service plan from one serer to another server
I want to add a DNS Failover automated service at my server, in order that when it goes down, the DNS Failover enters in action and it will redirect all the traffic to a second backup/failover server (hosted at a different location).
So in order to do this i have been looking around and i have found dnsmadeeasy, and i found their offer interesting.
But i still have some question and doubts on how a dns failover service works. So maybe someone can expalin it to me.
My question and doubts:
1 - If i use for example the DNS failover from dnsmadeeasy, i would have to buy 2 servers where one is the principal server and the other is a backup/failover server, so when my main server goes down, the dnsmadeeasy will detect in a automatic way, and redirect all the traffic to my backp/failover server. After my principal server goes back online, dnsmadeeasy will automatic redirect all the traffic to the principal server.
Im i correct? Or i understound wrong how this works?
2 - I use my own Nameservers and i want to continue to use my own nameservers. So if i use dnsmadeeasy services, i will be able to continue to use my nameservers as ns1/ns2 and then i will add the dnsmadeeasy ns3/ns4...as the following example: ns1.mynameservers.com ns2.mynameservers.com ns3.dnsmadeeasy.com ns4.dnsmadeeasy.com ns5.dnsmadeeasy.com ns6.dnsmadeeasy.com
I am planning to rent 2 managed dedicate servers with windows server 2008 and ms.sql 2008. which I will use one for web application and 1 for database server.
why people usually divide them into web and database server? is it for security or performance reason?
Which one you prefer to choose?
- 1 server for web and 1 server for database. or
- 1 server(web and database) and 1 server for failover
In making my plans for my web hosting business I plan to start, I've thought about a failover but wanted to know the best way to set it up.
I would ideally like to start with something like getting a small shared hosting solution (not reseller, just shared) at a reliable hosting company. On this hosting plan, I will host a simple 1 page site that says my company is having technical difficulties, reminds people that data is safe, and provides any contact information accessable (like e-mail/phone/etc...).
I already have DNS setup offsite.
I would like that if my server(s) go down for any reason, whatsoever, that DNS automatically starts serving the temp site.
so if my server suddenly goes down, [url]or [url]suddenly goes to the offsite shared hosting plan.
This would not provide customer site redundancy. This would only ensure that if anything happens to my service, customers have a page available that explains what is happening and has information to contact us to get more details.
We have 2 Cisco PIX 515's in a failover configuration (7.1 series IOS). We have available both a serial failover cable and adequate switch capacity to allow using either the serial or LAN-based approach. We have no current plans to locate the firewalls physically apart.
We have plenty of info on how to configure either, but no recommendation on which is the better choice.
I have a reseller account with a company that often has spontaneous DNS errors, and this, as you know looses us money. I also have a dedicated server running a panel exclusively via IIS. Is there a way I can set up my dedicated (Windows) box and my reseller account in a mirror way, so if one of them fails, the other will take over.
There are no file changes often, more database then anything. Is this possible? Main necessities are:
-Mail Servers
-MySQL Servers
-PHP5
-Cron jobs
(Or, if a shared hosting company can accommodate this somehow and is relatively inexpensive.
Basically have 2 hosting accounts at different providers...each set-up for the same domain name...and then somehow wtih DNS make it so if host #1 goes down traffic goes to host #2 (which would basically be a splash screen explaining that host #1 is down and will be back soon).
DNS isn't my strong point, but I do know you can do this with MX records...so if the first server fails it tries the next until it gets a working one or reaches the end of the list. I'd just like to do it is A records.
It wouldn't be as simple as setting the nameservers like this would it?
Would it use the host1 nameservers as long as they're online, and if not failover to the host2 nameservers? If so, great, but what if the host1 nameservers are online but the server itself is not.
Assume, a DNS failover service would change the DNS from server 1 to server 2 when the first one is offline. This will be changed quite fast, however isn't it true that some visitors still experience a downtime of a couple of hours (saying that the main server will be offline for 6 hours) as their ISP might not update DNS that frequently?
So basically, what is the added value of a failover service?
What happens if your primary nameserver goes down? Meaning that it goes completely off the net, not even denying dns requests but completely ignoring them.
I'd guess that resolvers would query the secondary NS after a specified timeout, but what is that timeout set to? Does it differ from ISP to ISP? How much of a slowdown are we looking at for end users?
I'm building the infrastructure for a live event site.
I have specced 8 app servers but I'm still unsure with regards to MSSQL.
The site functionality makes a single write to the DB with no reads. A simple option would be to have a MSSQL DB instance on each of the APP servers and simply collate the information after the event.
I've also been looking at the mirroring function of 2005 as it would mean we could get away with two MSSQL servers instead of 8.
Has anyone used MSSQL 2005 Mirroring? Is it easy to setup and maintain? Any caveats?
We want to provide failover service at our server using dnsmadeeasy service.
So we have already buy the "failover server" (running linux/plesk), and now we are looking to find the most adequate / best way in order to keep the 2 servers sincronized, in order that if "main server" goes down, all the traffic can be redirect to the "backup server" using dnsmadeeasy.
on scripts/software to keep 2 linux servers sincronized in a failover setup?
I'm working for a client who has an e-commerce site currently hosted on a shared hosting solution.
He is now looking for 100% uptime (as near as), so I have suggested that we get 2 VPS and use DNS monitoring to switch servers as required (from DNSmadeEasy).
This is all fine, and the websites files/images do not change often, so I can use rsync every so often to sync these. Not a problem.
What does change frequently is the mySQL db for the site.
I've been looking at MySQL replication, but this seems to be no good. If one server goes down, then the other one takes over, they don't automatically sync themselves after they come back up. It seems MySQL cluster is best, but this needs 3 servers and they all need to be on the same LAN.
I've read you can set MySQL replication to MASTER-MASTER so that it acts like a cluster, and resyncs itself as required.
Well my problem is, we have 2(ABC and XYZ company) internet connection and sometimes, the ABC connection is not active and we're disconnected for about 2-4 minutes interval to get the XYZ internet connection online.
My question is - Is there a software that will automatically set the XTZ internet connection to active? Something like if the ABC connection is down, the XYZ will supply internet connection right away.
I have been searching and searching for a solution. We are currently using one single vps to host some of our clients. We are finding more and more that we need to have some redundancy.
I have looked at using DNS failover using RSYNC/mysql replication etc with two servers, but just dont like the idea.
I have also looked at hosts like imountain etc that use h-sphere. I dont like this setup because services are split onto single machines. For example mail is done on one single server, therefor if that server is down, mail is down.
What I am looking for:
I am trying to stay in a budget of 150/month or less.
I would like to get one of the two options here:
option 1: two vps's or dedi's that technically act as one(a true cluster) then on top of that is OS and control panel and done. This solution doesnt allow for whole datacenter outages, or network issues.
option 2: Two geographically placed vps or dedi's that are somehow either load balanced, or failover.
Ultimately our goal is to have high uptime, but we dont really have much server load.
Basically failover is ok as loads are always low anyway, but if we are paying big $$ it would be nice to have it load balanced.
Please let me know if my expectations are way to high or my price is way too low. I need to find a solution here somehow and if I cant find anything will most likely just go with DNS failover.
