LongJump puts PaaS on-premise

LongJump puts PaaS on-premise

Summary: In a move that will horrify multi-tenancy purists but bring a smile to the face of many conventional ISVs desperate to launch SaaS offerings of their own, PaaS provider LongJump has today made its platform available as installable licensed software.


In a move that will horrify purists but bring a smile to the face of many conventional ISVs desperate to launch SaaS offerings of their own, platform-as-a-service provider LongJump is today making its platform available as customer-installable licensed software.

The move is designed to meet demand from enterprises and ISVs that want to run their own cloud development platform rather than entrust their fate (and a monthly per-seat subscription) to a cloud provider. But it will prove controversial with rival platform-as-a-service providers such as Salesforce.com and Zoho, whose argument has always been that the unique benefits of their multi-tenant platforms can only be realized by running on the cloud provider's own shared infrastructure.

LongJump takes a different view, and will offer its software for customers to install under either single-tenant or multi-tenant licenses, for an annual fee ranging from $60,000 to $240,000 per CPU. Customers will also be able to choose their preferred degree of multi-tenancy, for example whether to assign an individual database to each downstream customer, or how many different application instances they operate (allowing them, for example, to operate regional instances to cater for variations in data privacy requirements). LongJump is restricting support to its preferred operating environment of Red Hat Enterprise Linux and the mySQL database, and customers will have to keep pace with the company's quarterly release cycle, as it will only support the current and most recent version.

The most unique element of LongJump's proposition lets developers adopt a hybrid strategy, using LongJump's servers as a development or pilot platform while maintaining larger-scale production applications on their own in-house servers. This means an enterprise can standardize on LongJump as a custom development platform both for small-scale hosted PaaS applications and for larger in-house deployments. LongJump will also optionally host a customer's instance as a managed service, or if the customer prefers to host on Amazon, there's an EC2 AMI option that I've been told can be up and running within 30 minutes.

In my view, LongJump's hybrid offering plugs a gaping hole in the market today, both for ISVs and for enterprises. No conventional software vendor offers a fully equivalent hosted PaaS alternative to its on-premise platform (the closest parallel is startup Apprenda's multi-tenant SaaS platform for .NET applications). With the exception of Bungee Labs, no PaaS provider offers the option of a licensed installable version of its platform. And as LongJump's CEO Pankaj Malviya gleefully points out, none of the others can point to a five-year track record of serious customers running business applications on its platform. "There is only one other company [customers would consider]," he told me. "That is Salesforce.com. We are not afraid of them."

"Corporate IT is still not willing to put every bit of information they have into the public cloud," he said, adding that companies dislike having to pay the subscription Salesforce.com charges for every new user brought onto a custom application. As for ISVs, he said, "ISVs want a platform like Force.com — they love it — but they're not ready to lose control to Salesforce.com."

I am somewhat concerned however that many ISVs will leap on LongJump's solution as a means of adding a SaaS offering to their solution set without properly understanding what they're getting into — especially as having an on-premise version panders to the commonly stated belief that SaaS is 'just a deployment option.' Malviya himself explicitly rejects that view: "Software vendors who think this is just a delivery option don't understand the model," he told me. But LongJump won't be offering any SaaS enablement advice to ISVs beyond supporting the software.

"As far as go-to-market is concerned, we leave it to them," said Malviya, recommending that ISVs work with consultants such as Scio Consulting to help define their SaaS strategy. Helping ISVs be successful is not LongJump's role, he insisted. "They will have go-to-market challenges irrespective of which platform they're using," said Malviya. "Our core competence is to provide the platform.

"We see ourselves as a software platform company rather than a pure on-demand company," he added. "We want to appeal to companies that believe in our software but who might not want to believe in our infrastructure capabilities."

Topics: Salesforce.com, Cloud, Servers, Software

Phil Wainewright

About Phil Wainewright

Since 1998, Phil Wainewright has been a thought leader in cloud computing as a blogger, analyst and consultant.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • It's too bad

    As a purist, I agree with the article, none of
    the advantages of the Cloud can be gained by
    the creation "private clouds". These being:
    -Lower cost via economies of scale
    -Continuous innovation as a function of a
    greater ecosystem.
    -A constant "state of State of the Art" with
    which the traditional "on premise" data center
    just can't keep pace.

    But the issue plaguing LongJump that led to a
    management decision to move away from the pure
    PaaS offering and retreat to becoming an "Old
    Wrold Software Vendor" is one that all start-
    ups face. It is called capital. LongJump has
    reached a point where it simply cannot afford
    to keep pace with the likes of SFDC because its revenue stream hasn't allowed the company to
    attain the critical mass and therefore the very
    economies of scale for which the "As a Service"
    model is known.

    Couple this with the CEO's refusal to accept
    outside or venture funding and it was only a
    matter of time before LongJump was either
    forced to change its business model, accept
    third party funding or "PaaS Away" like so many
    of the other boot strapped ventures trying to
    profit off the latest buzz word...Cloud

    I wish LongJump good luck with its new entry
    into the Old World.
    Vladdy the Saas Guy
  • RE: LongJump puts PaaS on-premise

    I believe this is a long jump backward.
  • RE: LongJump puts PaaS on-premise

    Zoho used once the following terms: "Pragmatic, Non Dogmatic Approach to Winning Business".

    I guess, this is what Long Jump is doing.

    Might be a bad news to lose a PaaS evangelist, but are all the PaaS companies required to do so ?
    Yves Hiernaux
  • RE: LongJump puts PaaS on-premise

    Rollbase began offering it's PaaS as a licensed product in November of 2008 and has since signed a key customer which will not be announced until late 2009. Hosted plans are still available at roughly $50/user/mo (may decrease with volume)
    Matt Robinson
  • RE: LongJump puts PaaS on-premise

    Au contraire. They are doing what *Customers* are asking. This is just being wise rather than dogmatic.

    Now, they are offering options for their Customers to make a decision.

    If a Customer sees they have a way out of LongJump's datacenter, they might be willing to buy their offering.

    It will be interesting to see how many customer actually buy their on-premises version as opposed to having it "just in case" (LongJump becomes "CogHeade'd").
  • RE: LongJump puts PaaS on-premise

    From the viewpoint of someone trying to get enterprises to adopt PaaS as a concept, I would like to put forward a few thoughts arguing that on premises licensed "PaaS engines" is a necessity for certain categories of applications and customers. Furthermore, that this segment is likely to be relevant for many years still, and might actually contribute to driving the adoption of PaaS as well as cloud computing, rather than representing a step backwards.

    Starting with the prospects in question, let's look e.g. at a scenario where a global enterprise finds itself looking for a more lightweight platform than its standard ERP system -- that have proven to require huge investments and long lead times when business is changed to meet the frequent moves of the competition and market. The PaaS concept provides a clear alternative, and need to be considered even for the cases where the ERP system was previously dubbed the ultimate strategic solution. However.

    1. Dependability
    For the ERP system, the enterprise has strict SLAs with their service providers and vendors, ensuring system availability with 4-5 nines. Should the ERP-system not meet expected uptimes, the enterprise know that the "big-five"-level players they have contracted for hosting, network and software can pay the penalties stipulated in their contracts. Now, what about the typical PaaS offer? Maybe sufficient for gathering sales leads, but what about the core operations?
    2. Data security
    Not just the financial and health industries depend on data that is very sensitive -- any knowledge based industry does. The unwillingness to move sensitive data outside premises is a well know issue for the Paas industry.
    3. Vendor dependency
    Coghead's demise demonstrated the risk with PaaS quite clearly. Even if it is easy to export the data as such, it might not be as easy to rebuild a large application on a new platform, integrate it with the rest of the corporate IT landscape, and train and convince a thousand or so users that the new application is as good as the old one. Accomplishing this with a lead time of months in midst of a global recession? Tough luck! While this might not be a reel problem for the application storing your CD collection, it certainly is for the application driving a significant part of your corporations profit and loss statement.
    4. Economies of scale
    While economies of scale is clearly one of the pillars of PaaS, what if your customer is already operating a larger data center than you, as a PaaS vendor, will for the foreseeable future even with the wildest growth projections? They might not buy into the sales pitch stating that data center economies of scale is a key factor in the business case for them moving onboard your in the cloud PaaS offering. What if your prospective customer already have sourced global WAN links with QoS guarantees -- can the PaaS vendor compete with the customer in buying power when sourcing this?
    5. Performance
    Performance of business IT applications is a huge headache for many corporations today, and they spend plenty of resources to address and avoid this. Isn't it naive to believe that the PaaS vendor, lacking insights into business operations, will manage performance better and at lower cost than the enterprise? From day one?
    6. Integration
    Enterprises spend a fortune on integrating its business IT systems. Being realistic, how much of the technology used would be compatible with your in-the-cloud-PaaS offering? What is the sunken cost in these integration solutions and what are their expected lifetimes?
    7. Regional adoption patterns
    PaaS adoption is likely to take some time to reach around the globe, not to mention to reach beyond early enterprise adopters. While it does, on premises solutions might prove a very tempting springboard and market driver.

    To sum up:
    In my humble opinion, one can definitely argue that on premises PaaS engine installation is a good proposition for a large part of today's market. For most of the points above, an on premises approach provides a step towards a solution. It can also be argued that, eventually, the market might converge towards the purist's "PaaS in the cloud", and maybe be dominated by a few large players that have managed to achieve true economies of scale and "heavy weight" credibility.

    The road towards that however, might be longer and more winding than expected! And until then, there is no contradiction in going with either a cloud hosted or a on premises PaaS solution - a so called one size fits all approach is simply not likely to be attractive to all customers. But on the other hand, key benefits remain in place whether you go cloud hosted or on premises, e.g. short time-to-market compared to using traditional software solution (ERP).

    Fredrik Jonsson @ www.istools.com (A PaaS vendor offering both in the cloud or on premises solutions, depending on our customers needs.)
  • Hybrid model also available from Iceberg

    Just a quick note to say that Iceberg (www.geticeberg.com)
    has been providing its Platform service model for customers
    to install on their own servers since its launch last year.

    We have known from day 1 that some customers desperately
    needed to run their applications on their own systems in

    Plus the advent of coghead going under reminded people of
    the importance of avoiding vendor lock-in to a purely SaaS
    offering should the worst happen.

    Wayne Byrne