Skip to content

The Truth About Storm On Demand

As internet marketers the most important weapon in our arsenal is our web server.  A crappy server that’s constantly going down is a liability that WILL cost you money.  The typical options that you had in the past were shared server, VPS or dedicated server.  Any serious marketer that is BUYING traffic is probably on a dedicated server.  I have multiple hosting accounts including all of the previously mentioned.  One dedicated server in particular was constantly having issues.  The server housed an installation of Prosper202 as well as a few landing pages.  As soon as a huge influx of traffic would come in, the server would start not resolving pages.  Redirects also didn’t work so I would just lose money when the server went down.

At this point I was seriously considering buying a much more powerful server at another hosting company but Liquidweb convinced me to try their “Storm On Demand” cloud offering.  The salesperson that they had call me gave me a bunch of misinformation.  What I want to do is tell you how Storm On Demand cloud hosting REALLY works so that you can decide if it’s right for you:

What Is It?

Storm On Demand claims to be a cloud server provider.  There is debate on web hosting forums whether they are a TRUE cloud or not a true cloud.  Honestly, I don’t really care and neither should you.  What Storm On Demand allows you to do is provision a server (or instance) and get billed hourly for that server rather than a flat monthly fee.  You pay 3 fees with each server instance.

1) The Server Instance Itself – You can choose how much CPU power you want, how much ram you want and even what hardware is in your server using the “storm bare metal configurations.”  Pricing starts at $50 or $0.07 cents per hour for the lowest server specs and more as you add more resources.  If you only use the server instance for 15 days then that’s all your billed for.

2) Management/Software – If you want the typical Cpanel/WHM & Full Management that you are used to with a dedicated server it’s going to cost you $20 per month.  If you want to manage it yourself then no charge for this.

3) Bandwidth – Your typical dedicated server comes with a bandwidth allocation per month.  Storm servers do not come with ANY bandwidth included.  Bandwidth is charged in one of two ways.  You can either pay as you go for bandwidth at the rate of $0.15 cents per GB outgoing / $0.08 cents per GB incoming or you can pre-buy an allocation of bandwidth at cheaper rates.

This setup allows you to scale your server up and down at a whim (without contacting customer support).  What happens is you hit a button which scales the server either up or down with the specs you specify and then the StormOnDemand backend system “syncs” your data with another server instance that meets your specifications.  You are then billed at the rate of whatever the new server instance costs per hour.  Your resource use is billed similar to a utility company and is totaled up at the end of the month for your bill.   It also shows you what you are going to owe as you go.

For the lowest level Storm On Demand server you are looking at the following pricing (over the course of 1 month):

Server Instance: $50 (2GB Ram, 150GB Disc, 1 CPU)
Management/Cpanel: $20
Bandwidth: $0.15 cents per GB
Total: $70 + Bandwidth

If you get a more powerful instance setup it will probably cost you about what you pay for a dedicated server now but you have the ability to scale up and down much more easily.

Why Would You Want This?

Depending on the type of traffic you buy, you may experience that you get huge volumes of traffic in spurts.  Maybe even only certain days of the week or even certain months.  Storm On Demand allows you to cut the waste and make sure you always have the resources you need.  Let’s say you don’t run any traffic on the weekend.  You can scale down from a super server on Friday night and scale back up on Monday morning to save some cash.  The possibilities are really endless (especially if you can write some scripts because they have an API).


  • Scale your server up and down to account for traffic spikes and dips
  • Pay about the same as a dedicated server for a better server only when you need it
  • Deploy additional servers and load balancers if you need them
  • Store images of your server so that you can restore to a previous state any time you need to
  • Super fast
  • Flexible billing
  • Works exactly as your dedicated server does with the same Cpanel setup
  • Anything you can run on a dedicated setup you can run on this

Disadvantages & Needed Features

  • In my experience the resizing process takes 35-60 minutes with a few minutes of ACTUAL SERVER DOWNTIME towards the end of the process.  The process ends with a reboot of the server which completely takes it offline.  The sales guy DID NOT tell me this, so I am telling you.  If Storm On Demand could figure out how to complete this process without ever actually causing downtime, it would be way better.  If the resizing process could take closer to 5 minutes that would also be helpful.
  • There is no scheduling.  Let’s say every night at 8pm you want to re-provision your server resources.  You can’t without logging in or writing some kind of script.
  • Bandwidth on the pay-as-you-go plan is too expensive.  The whole point of cloud based servers is to be able to adjust with the flow of traffic.  Requiring you to pre-buy a package of bandwidth in order to get a better price kind of defeats the purpose of the flexibility.
  • There are no email alerts to tell you when you have reached a certain bandwidth limit.  This would be helpful so that you can monitor what your bandwidth is costing you if your not the type of person who wants to login to their server control panel often.  Also, i’m not sure how Storm On Demand would catch an issue like a DDOS that would suck your bandwidth and cause you to get a super costly bill at the end of the month.  An email alert at a predetermined GB usage would help.
  • A way to setup server rules would be helpful.  For instance, “If my server reaches 80% of it’s resources, resize to the next level automatically.”  With the current system you have to monitor this yourself and choose the appropriate settings.

Overall, I suggest you giving Storm On Demand a try.  I think this style of offering is the future of web hosting.  My new setup i’s the smoothest i’ve ever seen my Prosper202 run, my landing pages are serving up faster and redirects are noticeably quicker as well.

Have you tried Storm On Demand or any other Cloud based servers?  Share your thoughts…

Note:  No compensation was received for writing this post. All points made are Ad Hustler’s opinion.

Ad Hustler | Subscribe To Ad Hustler

Published inDoing Business


  1. I recently switched everything over from dedicated servers to storm and I gotta say it is amazing.

  2. Jessica Jessica

    Yeah, the part I hate most is the resizing part. I’ve asked if there was any way it could be optimized to be shorter, but they haven’t come up with anything. I just have my prosper install on it and my bandwidth doesn’t come out to be too expensive. The most I’ve paid is $5-9.

    I still use a dedicated server for actual landing pages and such since cloud bandwidth is more expensive.

    Another CON that had me move over to Bare Metal Storm was that recently it seems like there is some older hardware that they’re using for the normal VPS. Performance wasn’t as great as it was before. Moving to bare metal is pretty much the same as resizing servers between the normal VPSs. The sales guy told me a bunch of erroneous stuff about it though when I was first asking about it.

    Overall, I like it a lot.

  3. The flex option (scale up/back) sounds awesome.

    I’m going to take a look at these, thanks Brandon!

  4. After reading this post I’m considering making a move over to Storm. I recently shut down a dedicated server and placed a bunch of sites on different hosting packages. My money sites have different requirements than some of my other sites that may be heavy with content but lacking traffic. (search driven)

    I wish I knew about this before! Knowing that you have the bandwidth to push serious traffic is an insurance policy worth taking. Having the ability to scale up and down on the weekends is a great feature.

    Thank you for sharing this. Much appreciated.

  5. Try out the Rackspace cloud offering. I was floored when I tried it – Price-wise they have offerings that scale lower than $.07 an hour down to 1.5 cents and hour and as high up as 15GB of ram for 96 cents an hour. Its the same deal with the bandwidth – a little on the pricey side. You can also instantiate Windows servers – It was pretty cool to fire up a Windows server, connect via remote terminal and surf remotely (same via X11 and Linux, though you gotta go and get X running yourself).

    Oh.. And they don’t tell you that if oyu want your server to stay in their system and not just get deleted, you’ll have to pay for its disk storage. Either way, it works out to a reasonable cost… Check them out. This is one area where I personally ain’t taking any more chances.

  6. Makes since if you get big traffic spikes or your ads get picked up on big sites.

  7. I haven’t tried cloud hosting yet, but it seems to be the new wave of the future, especially with all the flexibility it seems to have.

    Mind me asking though, couldn’t you just write a Cron Job (or maybe I’m understanding this wrong, sorry) to revision the server every evening at a pre-set time?

  8. @Dustin – I don’t think it’s as easy as writing a cron job because the re-provisioning doesn’t take place in WHM or Cpanel. It takes place behind a password protected control panel on StormOnDemand’s site.

  9. Justin Justin

    Haven’t used this either yet but would it not be possible to code a bot that logs in every 10 minutes and checks resource usage. If your nearing limits it could either then mail/text you or just automatically request a scale up?

  10. brian brian

    I’ve been using Storm on Demand ever since someone recommended it to me and I’m loving it. Your pros/cons list is pretty accurate though, and the bandwidth can be a serious cost if you’re running massive traffic. But their support has been great on their fully managed servers…

  11. @Justin – You probably could write a script to do that. It would just be nice if it were an included feature in StormOnDemand.

  12. We always get questions on why PPV traffic is dropping clicks… Funneling 1mil plus through your prosper database on shared hosting is detrimental. Get off the shared! Nice insight on storm, I appreciate it!

  13. Looks nice. I’ve been on a few dedi’s and VPS’s with liquidweb and I think I might have to try to consolidate it all to one using this.

    Thanks for the write up!

  14. @TrackingLex – Yea Storm is perfect for Prosper202. You should refer people to this post because i would have loved to read this when I was trying to figure out how all of this works

    @Volk – I haven’t even needed the bare metal yet. For some reason even their lowest level plan runs much much better then a server with the same specs so far i’ve been switching between the 2GB, 4GB & 8GB (Mostly the 4GB). I haven’t really been blasting the server as much as usual though because i’m still kind of testing. It has however handled some pretty heavy loads quite nicely.

  15. […] The Truth about Storm onDemand – This post convinced me to move all my 3 VPS servers and 1 dedicated server onto a one scalable web hosting solution. Once I move it all… I will be sure to post how I like it, etc. […]

  16. Tom Tom

    How does this compare with Amazon’s Elastic Computing offering, which will scale automatically to up to thousands of servers (and lets you specify what country(ies) you want the servers in)?

  17. @Tom – It seems like EC2’s pricing is pretty similar, it’s harder to understand their offerings and I don’t believe that it functions exactly like your standard dedicated with cpanel, WHM and all that jazz. I’m no expert in cloud hosting though so if you decide to try it post here and let us know the differences.

  18. Dang…. after doing so much research here it would make your head melt.

    Spending too much with many other hosting providers, 9 dedicated servers at hostgator and something at rackspace, and the planet.

    We too looked to the cloud.

    I checked out Storm, though in the end went for a different offering Ad Hustler.

    We used

    main reasons….

    1. bandwidth is bundled in with it, and more bandwidth is substantially cheaper.

    2. heaps of pre-configured images for doing some fairly advanced things like mysql clustering.

    3. cheap cpanel/whm addon

    4. Other integrations like server stats monitoring, using litespeed vs apache very easy to integrate.

    So far, very impressed and liking the whole, use what you need.

    As opposed too…. rent a server that will let you expand because you will need it later, but only use 10% on an ongoing basis.


  19. StormOnDemand is the awesome. Mucho faster.

  20. storm?? lol wait for a ddos attack and the automatic charge to your credit card…

  21. @John – I do agree that this is an issue – Their needs to be some kind of way to set a limit to your bandwidth usage.

  22. John brought up a good point here in the comments regarding what happens in there is a DDOS attack against your Storm On Demand Server. I sent Liquidweb a message on Twitter asking:

    @StormOnDemand Question from the blog: What do you guys do to prevent DDOS attacks and if one happens are we responsible to pay bandwidth?

    Benny @ LiquidWeb was kind enough to respond via email:


    Hey there,

    The short answer is: It depends on the level of attack. If you get at
    10M DDoS aimed at a site on your server, and you configure your
    software to handle the load (and/or we help) to keep up with the
    traffic, then that’s what we do. At that point, yes, you would be
    responsible for the cost of the traffic, but we would probably work
    with you to help prevent it from being exorbitant.

    If you get 4Gs of traffic aimed at your site, in order to keep our
    network stable we would be forced to null route the IP, which
    basically means all traffic to or from that IP (not your entire
    server, just that IP address) would be dropped.

    Let me know if you have any other questions.

    benny Crampton
    SysAdmin. Evangelist.
    Liquid Web, INC. / StormOnDemand
    @liquidweb / @stormondemand


    I also asked him why there are no alerts when you exceed a certain amount of bandwidth to at least warn you of a situation like this. He responded that they are actually developing an alert system right now and should be out in a release or two.

    It’s nice to see that StormOnDemand is working to make an already good product even better.

    I’m still using Storm On Demand myself, and I still like it as of writing this.

  23. Volomike Volomike

    Some affiliate competitors may find your sites and flood them, or someone who hates your kinds of sites may flood them, and there goes your credit cards. Unless you have a crack sysop team who know their stuff and can configure routers and Linux, and be available 24x7x365, these cloud-like and on-demand sites are really not going to help you cut costs. My clients have tried them, and they have often failed.

    So don’t believe the hype. Stick with regular dedicated or regular VPS (several of them), and get some great sysops.

  24. Switching to the cloud can be hard but the pros far outweigh the cons. Give it a try, there are several companies out there that can help you. Storm on Demand will give you a coupon for a couple free months if you call them up on the phone. Tell them that you aren’t interested, and then wait for them to call back!

  25. Jon L. Jon L.

    Just an FYI, Storm now has an API in BETA. They signed me up for it just this week, so resizing should be possible via the API now (cron job, or monitored performance, etc)

  26. Vu Vu

    I worked with rackspace cloud server for the past 5 months. They have overing under $0.07. I have a few server instances with them and I up size and down size daily.

    The cloud control back end is very slow. Online they don’t have live detail of all the sizes that I switch and their charges. I have to keep track of that myself to make sure they are charging me correctly.

    Also, I’ve had instances were I go up size and cannot down size, so they are raked up the charges for that. I had to do a ticket to get help down sizing. I got stuck with up size and cannot downsize for over 3 days and still not had anyone responsed to my ticket. The chat help was clueless.

    I did some research on the prices, rackspace is over 2.5 more pricier than storm on demand on comparable plans.

    I am planning to move my production instances to Storm to get more juice for the money. I keep my development boxes on rackspace because I can keep them at the 256MB plan when I don’t work on them.

  27. Jon L. Jon L.

    I concur with Vu about Rackspace vs Storm. Rackspace works well for my private server (512mb instance), and is cheaper than Storm could go (I pay <$30/mo at Rackspace), but if I ever size up, it'll be to Storm.

    For my employer, we're actually starting to transition from SoftLayer to Storm, as SoftLayer's cloud performance is abysmal and more expensive than Storm.

  28. Dude Dude

    Another con with VPS is that you are still sharing resources with others on the base system. One place where you can really take a hit is disk I/O if you have an app with lots of traffic which hits the database (WordPress.) For most people, the DB will be the bottleneck, if you are running a VPS then this bottleneck is worse. The bare metal servers are also using XEN software (same as their VPS plans) but the difference is the bare metals have only you on one server. Though you aren’t sharing resources, you might still be taking a hit on overhead. I’m not sure how the I/O compares between the two plans, but if the above issues are a concern for you, then you might as well go with a traditional dedicated server.

  29. Mike Mike

    I’ve been a Storm user for about 5 months. Many of the items noted above are 100% accurate ie: downtime for resizing etc. Be prepared for when you start up and are configuring your instance, they will likely block your ip, and you will have to contact tech support to whitelist it.

    Yesterday I discovered a big problem, when according to the Storm proprietary panel my instance used 6TB (yes that’s right) of bandwidth in ONE WEEK, adding about $800 to my bill. Check WHM and Bandmin, and according to those my bandwidth was in normal range (about 350 GB, and well within my package of 500GB)

    Needless to say I flipped. After tech support passed me to sales, and sales passed me back to tech support, I’m told sales would handle it, and I would hear from them today.

    But of course I am very concerned. I was literally about to pull the trigger on a new instance (trying their new cloud/dedicated hybrid bare metal server), but now I am worried.

    Running the servers is just ONE part of my job (unfortunately), and Storm does make it easier. However, when your on the bleeding edge of technology, be prepared to bleed a little, and I am certain it will get better over time. No technology is perfect, especially new ones, and this developments in cloud computing certainly offer a great benefit to overworked/underpaid geeks like me.

  30. chaos1 chaos1

    I have been using Storm On Demand since September 2010 on the 2GB package. What can I say? Deals! Substantial load times since we moved our major sites off our previous dedicated server with LiquidWeb. We are now running a network of 154 sites on a SoD instance. Today we decided to scale it to 4GB package. The last week we’ve been hitting the ceiling and dipping into swap.

    After looking at dashboard statistics, web stats, analytics, optimizing our databases and our load times we’re happy to see the growth. May 2011, we received 474,288 page requests on our major site, excluding a couple other cPanels that were brought onto the storm server 2 weeks ago.

    I must say, we’ve hit OOM errors in the last week. We optimized Apache, MySQL and looked into what we could do for caching. We may potentially spend more time in the back end (v.s. optimizing our theme load times). Now that page_speed is stable, we might start poking LW techs to see what there opinion of it is. I’d love to implement page_speed in the near future. Has anyone fired it up on their instance yet?

    Thanks Op for this write up! I found this post while Googling around looking for others experience with SoD. You should update this with 2011 edits!

  31. Ed Ed

    This is one of the few, rare reviews on StormOnDemand.

    It would be nice if you could do a review follow up, they have recently reduce their price on transfer, and i think they have new API for scheduling and Auto Expand etc.

  32. I became unhappy with storm on demand and moved what was hosted there to SolarVPS

  33. Ed Ed

    Interesting, could you do a short blog about why you moved, what happened? and may be a few words on SolarVPS?

    I believe a lot of surfers are still using your review as guidelines, ( Since it is the first few to come up with google )

  34. Michael Michael

    I just found this out myself. I was preparing for launch in earlier spring and just asked about that ‘spike’ process and was told unlike say RackSpace CloudSite it is not automaticaly provisioned/scaled, that it requires both manual monitoring, intervenion and reboot.

    I love Storm/Liquid web for their customer service which I think is unparalleled but this is not true cloud service as I understand it.

  35. Jan Jan

    In the last two days Storm experienced large DDOS attack. Unfortunately some of our IP addresses were affected.

    The issue Was solved on Storm side closing our ip address and refusing all the connections incoming to port 80 for the web server, but they didn’t tell us nothing.

    We came to the truth after two days of long phone calls on 3rd level support.

  36. I have 3 accounts with SoD and have been with them for over 3 years. Their tech support is phenomenal and never seems to amaze me. I’ve been through many hosts and horror stories, but I haven’t had any issues with these guys. They are definitely worth looking into. I am sorry to hear @Jan had outage issues with DDoS attacks. There seems to be no escape from these things. Amazon had its own issues this year, with big sites going down and then godaddy’s DNS messing up causing a large portion of the interwebs do be unreachable.

    These things are inevitable but and when the poop hits the fan it’s all about the companies’ support and Storm on Demand has amazing support.

  37. I use or (same thing) and they are the absolute best, quickest amazing host I have EVER had. I have a problem – BAM! it’s fixed. They migrated my whole website to them for free. Databases, the whole shebang, then stuck around to make sure it all ran correctly. They are on heck of a hosting company! Many of these posts are old and I don’t know what happened between then and now but I sure am glad. I lead in 483 ms with a cdn. Crazy good hosting.

  38. Thanks for another wonderful post. The place else
    may anyone get that type of info in such a perfect way
    of writing? I’ve a presentation next week, and I am at the look for such information.

  39. Alan Alan

    If you want to speed up server resizes, move your data to a block storage disk. Resizing involves 1 reboot, a copy of all data from the main disk from one server to another, and another reboot. Make that middle step shorter by moving either all of your non-configuration data or the largest files onto a block storage device. Since that is a network disk, it won’t be copied and the move will be much faster.

Leave a Reply

Your email address will not be published. Required fields are marked *