Hello,

Today I have compared Mosso's Cloud Sites offer to Cartika's HAL Reseller offer and both have drawbacks. Mosso's control panel doesn't look customizable for the end-user, they don't accept Paypal and don't provide a "Sign Up" application for new customers to join. Cartika's price per GB is twice as much as Mosso, and I wish their uptime were better. What do you think is best? :)
Thanks,

Generally, Cartika's HAL should be considered more reliable. If you check status.mosso.com and the announcements of Cartika's forum, you will see... Mosso seems to have issues EVERY day...

As for Cartika's uptime - last month they had physical transfer of the servers in their own datacenter suite inside colo4dallas - they could pause uptime stats and be close 99.99% for that month, but they prefered not to do so...

Some other point also to consider: Cartika is using H-Sphere and if you wish to move on the future this would be quite easy (ask Cartika to do the export and your new provider will migrate your sites), while Mosso is using proprieatary cp - if you wish to move, then you have to transfer each site manually.

But both choices should be rock solid - both companies are famous for their support and reliable service.

Good luck!

Thanks!

-----------

All HSphere providers have a clustered cloud-like hosting environment :) Incase you are looking for more options

Don't go with mosso -- trust me, save your time. I currently have the OpenVPN website hosted at mosso and they have been terrible! Back to using our own cluster...so much for saving some money.

@Yash-JH: Thanks, I'll take a look at JodoHost.

@RackPoint-Morgan: Could you give more details about the problems you had with Mosso?

I've seen issues reported by customers in regards to Mosso, but not for Cartika. From the perception I get, Cartika is very customer service oriented which is just as important as the providing the service itself..This can be attributed to Mosso being owned by Rackspace, which I would say their dedicated server solutions is what they have their eyes on the most instead of Mosso. If you haven't already done so, send them both a few pre-sales questions..

All HSphere providers have a clustered cloud-like hosting environment

No, HAL is not the typical clustered service. HAL at Cartika is a completely load balanced, redundant, clustered system, which is much different than standard clustered offerings.

All HSphere providers have a clustered cloud-like hosting environment :) Incase you are looking for more options

Yash, you should know better then this and should know me better then this.. I would not advertise an hsphere standard cluster as a cloud, etc..

we offer both standard clustered hosting

http://www.cartikahosting.com/clustered-hsphere-hosting/

and the HAL, cloud hosting

http://www.cartikahosting.com/HAL/

A standard hsphere cluster is nowhere near the same as the HAL offering


Cartika's price per GB is twice as much as Mosso,

Hello Mike,

This isnt exactly true. Mosso also charges for CPU usage, where we do not. So, our cost per GB usually works out to roughly the same or less. Now, with Mosso, if you have a poorly written script, they will hammer you (rightfully so) with CPU usage fees. If you have a poorly written script with us, we will work with you to fix it or you will need to move it off our system...

and I wish their uptime were better.

Our uptime is fine.. we did have a physical, scheduled move last month which really did hit our uptime stats.. (cant ever see that happening again)

after that though, the primary issue is the 3rd party monitoring service we are using (pingdom) - they measure in 2 mins increments - so, a 99.99% uptime SLA can only accommodate a 2 mins outage.. if we have a 1 second outage, or if 1 node fails and the monitoring session is tied to that node in a sticky session - the monitoring service will automatically flag this as a 2 min outage - when in reality - it may have been a 1 second outage, or no outage at all while requests are sent to other nodes.. (sticky sessions are required in load balancing shared servers, as many client sites rely on sessions - what this mean is if the node they are tied to fails, users will need to let their session expire or hard refresh to force a new session onto a new server - this is simply how load balancing must work - this is not considered an outage, this is however considered a session loss for any user with a session tied to a server that went down)

We will either need to find a more granual 3rd party monitoring service (ie something that can monitor down to the second), or we will need to change our uptime SLA for HAL - as it is getting too difficult to maintain a 99.99% uptime metric with 2 min monitors - actually - its pretty well impossible.. we are happy to use our own monitoring system, but, clients do not always trust those numbers (I can understand this) - so, if we must use 3rd party monitoring systems, and if they only have the capability to monitor down to 1 or 2 mins intervals - then we cannot possibly SLA a 99.99% uptime as it is mission impossible with 1-2 mins monitoring intervals to achieve this...

A standard hsphere cluster is nowhere near the same as the HAL offering


Andrew, I simply stated a Hsphere Reseller solution (properly done) offers a clustered, cloud-like environment for Reseller Hosting. I never said your service was worse or better. You can claim your HAL offerings are way better than the competition. All I am saying is you still use the HSphere-based clustered control panel to offer what you do offer!

I don't know how good your service is, I am sure it's great. Since you brought our service into question, I will also state that we too offer a clustered, cloud based Reseller Hosting solution (based on Hsphere) at a different price point that has consistently delivered great uptime and performance.

No, HAL is not the typical clustered service. HAL at Cartika is a completely load balanced, redundant, clustered system, which is much different than standard clustered offerings.

Thank you all. I'm just a little puzzled...I read on Cartika's site "HAL Cloud Hosting", and here I hear that HAL and cloud hosting are "much different". Could someone clarify, please :confused:

Thanks again.

Yeah, this things confuse me a lot. It would great if somebody clear the air on the differences

The best way to compare is to contact the providers themselves, as the term is used too generally.

Andrew, I simply stated a Hsphere Reseller solution (properly done) offers a clustered, cloud-like environment for Reseller Hosting.

Yash, this is simply confusing the consumer.. A standard hsphere services cluster is not the same as utility hosting.. (call it a cloud or a grid or whatever you like)

I never said your service was worse or better. You can claim your HAL offerings are way better than the competition.

Yash, this is quite an accussation, I am a little disappointed. I will assume this is just a simple misunderstanding and move on - fair enough?

Can you actually please show me where I said our HAL system was "better" then the competition? or heck, where I said it was "better" then a standard hsphere cluster?

All I said was that is is "different" then a standard hsphere reseller cluster

In fact, I do not myself believe one is "better" then the other - which is clearly why we offer both. However, I do believe that each solution has its place and that different customers with different requirements will and do opt for different offerings based on their requirements...

All I am saying is you still use the HSphere-based clustered control panel to offer what you do offer!

Yash, a control panel and architecture are 2 separate things.. As you are well aware, a control panel sits on top of architecture.. for example - you can install hsphere on:

1) a single server, or
2) you can run hsphere on several servers - each running a different service (think your reseller accounts and our standard reseller accounts), or,
3) in the case of our HAL system, you can run hsphere over arrays of load balanced servers/vps with centralized storage..

As you can pretty easily see, the control panel is the same in each instance, but, the architecture dramatically varies..

actually, the very basis of cloud technology is the separation of application layers from architecture - and utilizing those application layers to spread sites and services across a theoretical infinite number of nodes (virtual or physical) as well as scale storage requirements ondemand from a few GBs to 1000s of TB's if required...

I don't know how good your service is, I am sure it's great.

thank you for the kind words

Since you brought our service into question,

I actually never did bring your service into question - As I have said on these forums many many times - Jodohost is a top notch hsphere provider - I always recommend them when asked..

I will also state that we too offer a clustered, cloud based Reseller Hosting solution (based on Hsphere) at a different price point that has consistently delivered great uptime and performance.

Yash, this is really amusing at this point.. you are actually purposefully causing confusion - as can be seen by the posts in this thread - and it appears you are doing it on purpose..

Again, I am hoping and expecting this to be a simple misunderstanding - maybe you have not look at our HAL offering and maybe you feel we are just rebranding a standard hsphere offering? in which case, no big deal at all...

Yes, you do offer a clustered based reseller hosting package. It is exactly similar to our clustered based reseller hosting package - and yes, you have a different price point then us (are cheaper) and on the whole, seem to offer a great service..

Having said this, as far as I am aware, you do not offer a load balanced, scalable offering with centralized storage do you? a solution where you can scale a user beyond the limitations of a single server? call this what you will - a cloud, HAL, a grid, etc, etc, etc - but, it is what Mosso, Mediatemple, Amazon, our HAL offering, etc - all offer - your hsphere services cluster is not the same thing - and saying it is the way you are is not only confusing myself - but, is also confusing consumers... simply running hsphere does not determine your architecture as far as I am aware...

Thank you all. I'm just a little puzzled...I read on Cartika's site "HAL Cloud Hosting", and here I hear that HAL and cloud hosting are "much different". Could someone clarify, please :confused:

Thanks again.

Hello Mike,

clouds, grids, HAL, whatever - all represent a certain architecture...

Yash has caused some confusion here - and I am still hopefull its an honest mistake on his part, vs competitive type of action...

It is actually pretty straight forward...

A standard hsphere services cluster (which we have been offering and selling for years - as has Yash at Jodohost) - is a group of servers, each running a specific service

For example

Server1 = web
Server2 = mysql
Server3 = email
Server4 = DNS1
Server5 = DNS2
Server6 = Control Panel

A services cluster like the one described is what Yash runs over at Jodohost and is what Cartika runs as our standard clustered hosting offering. This technology offered some very large benefits over the previous legacy systems (ie single server cpanel setups, etc) - and companies like Jodohost and Cartika grew very quickly and did very well because of this competitive advantage...

Over the last few years, the trend in the industry has been moving towards utility types of models, scalable solutions, ondemand computing, redundancy, etc, etc, etc.. call this whatever you like - some call it grids, some clouds, etc, etc, etc..

In a "cloud" type of setup, you utilize a physical and/or virtual nodes to create a redundant, scalable environment. Typically, load balancers and SANs are added to the infrastructure - but, they are not always built this way...

In a "cloud" or utility based model, the architecture would look a little different... you cannot speak "servers" anymore - you need to speak arrays or groups or "clouds" as it were...

so

array 1 = web service (can be spread over multipe servers or VPS instances and typically connected to SANs)
array 2 = mysql service (can be spread over multiple servers or VPS instances and may be connected to SANs)
array 3 = eMail service (can be spread over multiple servers or VPS instances and typically is connected to SANs)

now, some still yet do this differently - where they will have the following

VPS instance - with email, DB and web - replicated over 2-10 identical VPS instances with requests load balancing between the instances..

the distinguishing feature of these systems is the ability to add more resources ondemand.. if a site gets massive traffic bursts, a user is able to pay for the bursting, when they need it.. so, if they need the equivalent of a $10 hosting account 11 months a year, but, need 3 dedicated server worth of resources for 1 month of the year - they just pay for their usage when they use it.. in legacy hosting systems, customers would need to migrate to dedicated environments during their peak periods, or they would need to pay to maintain dedicated environments all year, even when they didnt need them..

Having said all of this, these cloud systems are new. I would not say the technology is perfect yet.. although the technology is designed to remove single points of failure by introducing redundancy - and the technology is designed to accommodate large bursts in a utility type of format - the technology does introduce certain complexities. For example, you are now dealing with SAN systems which themselves do have some limitations. Often times VPS technology is introduced into these systems, which adds yet another application layer on top of the physical architecture - which then typically has yet another application layer (ie hsphere or Mossos control panel, or some are even trying to use cpanel for this, etc) - and well - the end result is an added layer of complexity..

I am very proud of our HAL/cloud system, but, I think it has been successful because we are careful to only direct specific types of customers there.. for the vast majority of hosting customers, a more traditional hosting account is probably still the best fit.. which is exactly why we still offer our standard clustered hosting accounts and why we do not expect them to be going anywhere anytime soon..

hopefully this helps ...

High Availability, Cloud and Clustering are not really the same thing - they can be used together, they can be used in the same context... but they're different.

As far as I am aware, H-Sphere is merely 'clustered'. I.e. It has separate, multiple machines for different uses. If one machine fails, the rest of the cluster continues working as normal. Of course, you could host your H-Sphere cluster on a series of Virtual Machines on a cloud hosting platform to help mitigate against hardware loss, but I'd be unsure of the performance...

High Availability can apply to cloud hosting, it can apply to load-balanced hosting, it can apply to automatic failover... it's just a name for a collection of methodologies of making your website... high availability :p

As far as I am aware, H-Sphere is merely 'clustered'. I.e. It has separate, multiple machines for different uses. If one machine fails, the rest of the cluster continues working as normal.

actually, the standard hsphere install does not do this.. if a server or service fails, that server or service is down until it is brought back up...

Of course, you could host your H-Sphere cluster on a series of Virtual Machines on a cloud hosting platform to help mitigate against hardware loss, but I'd be unsure of the performance...

this is what our HAL cloud offering is - a series of physical and virtual machines utilized to provide redundancy in hardware and services.. more importantly then redundancy though, it is used to provide scalability beyound the limitations of a single, physical server

High Availability can apply to cloud hosting, it can apply to load-balanced hosting, it can apply to automatic failover... it's just a name for a collection of methodologies of making your website... high availability :p

not really, high availability is one piece of a utility type of model.. whether you want to call the utility model a grid, a cloud, HAL, whatever... the point is the movement towards utility hosting.. and again, High Availability is just one piece of this puzzle..

not really, high availability is one piece of a utility type of model..


High availability is a system design protocol and associated implementation that ensures a certain absolute degree of operational continuity during a given measurement period.

I was just saying that High Availability doesn't refer to any specific technologies... :)

I was just saying that High Availability doesn't refer to any specific technologies... :)

Hi Darren - I completely agree with you... just trying to clarify what we offer is a little different then that..

truth me told, we did not want to market our system as a "cloud" - but, since we already offer a pretty highly available clustered setup - it caused too much confusion - we needed to brand this new system as something "different" that would better enable the consumer to understand the differences in our offerings - unfortunately or fortunately - "cloud" seems to be the accepted terminology amongst consumers for this type of offering..

Andrew, I believe you seek to cause confusion by relating your "HAL" services with a true grid platform like Mosso.

Your HAL service is a simple extension of HSphere. Hsphere already has the ability for SAN storage (which is nothing new in the industry) and load-balancers. So if you are saying you have innovated and offered something drastically different from an HSphere Cluster, than I beg to differ. Also, your load-balancing is clearly limited to Linux hosting, which is an HSphere limitiation :)

I am very well educated with the terms load-balancing, cloud networks and their capabilities. I have worked on a very large corporate cloud network as a Software Engineer (at Cisco Systems) and have seen load-balancers in action!

To begin with, I have never seen the utility of load-balancers in a classic, shared hosting scenario (whatever the load per server is). In class shared hosting, the point of failure is usually NOT the hardware (and you know this). We run quality, redundant hardware at JodoHost (such as redundant disk arrays, servers with redundant parts and a network with redundant routers, etc). I am sure you run a similar, well maintain hardware/metwork :) Coming to my point, with shared hosting (where websites are hosting by different companies on the same server), the points of failure are largely software related. And for software-related issues, an experienced Server administration team, working 24x7 is required to maintain good uptime. Proactive measuring of the websites, on these servers help! That's what we do, and I'm sure you do too.

So I'd claim, that load-balanced web servers (which you dont offer on Windows) will not improve uptime significantly. We've maintained 99.9%+ uptime on our servers, with simply redundant hardware and a redundant network. Load-balancing best works when a SINGLE application has to be scaled over a several servers.. Load-balancing was never really designed/intended for several or 50-100 independent domains being load-balanced on a server, as its simply not as effective!

Additionally, you claim you offer a truly scalable solution. The only people that can claim to do so, are Mosso and true-grid providers which are very much larger than Cartika, have alot more man-power and have a huge technological leap over your services. Scalable storage means nothing these days (in a shared hosting environment again) when you can't scale the processing power and RAM on your servers infinitely too!

Very simply Andrew, your services are not fully scalable and are in fact no more scalable than any HSphere cluster's services. And i would argue that load-balancing doesn't add any significant uptime advantage :) Our HSphere cluster is just as scalable, and we too have consistently delivered great uptime and service at our price point

Wow, simply amazed by your replies lately yash. First to interject yourself into this thread when no one brought up your company and make a comment such as in post #4 of this thread that "All HSphere providers have a clustered cloud-like hosting environment Incase you are looking for more options" is not only rude and deceiving, but it's borderline solicitation in these forums.

Then once the differences are made clear between completely load balanced and redundant services…you still come back and talk "smack". Sorry sounds like sour grapes to me as an outsider. Now I will say that I am a bit biased when it comes to Cartika as I have been a long time user of their services, but I do know the difference between their HAL offerings and a standard H-Sphere cluster…and you don't have such a setup. I'm not saying you couldn't, but you don't, and they are not the same, so lets stop deceiving people by saying they are the same. Whether you see the value in it is beside the point and to state things like "I have worked on a very large corporate cloud network as a Software Engineer (at Cisco Systems) and have seen load-balancers in action!" means absolutely nothing to me, just as me saying I "used" to make 80K a year from one website that took about 15 minutes of work a week means nothing to you. We all know there are people working at McDonalds that "used" to do other things.

The rest of your post above is again "sour grapes" in my opinion and sure doesn't reflect well on you or your company when you have to try to tear down another company to try to make yourself look better. Especially when you have to compare them to another company instead of your own to do so. Sorry, just an outside opinion looking in. I have noticed in another post recently where there was deception on your part…not looking well on you. I'm not trying to get into a pissing match..but you haven't been doing you or your company any good in threads with comments like this.

tcstatic, I am glad you are a loyal Cartika customer :)
Your post is nothing but a personal attack on me, which you are free to do of course.
However, my post is not an attack, rather objective debate.

Andrew, I believe you seek to cause confusion by relating your "HAL" services with a true grid platform like Mosso.

actually Yash - our architecture is pretty close to the MT grid and the Mosso cloud.. funny enough, they also have a control panel on top of their architecture..

Your HAL service is a simple extension of HSphere. Hsphere already has the ability for SAN storage (which is nothing new in the industry) and load-balancers.

sure hsphere supports load balancers and SAN's.. last I checked though, hsphere does not support mysql clustering, mysql farms - nor does it support Xenserver and virtualization

http://citrix.com/English/ps2/products/product.asp?contentID=683148

and I quote:

"Citrix XenServer™ is the only enterprise-class, cloud-proven server virtualization platform that delivers the critical features of live migration and centralized multi-server management..."

We are actually in the process of integrating this within hsphere so resellers can allocate and sell Xenserver based clouds running hsphere, right from their control panel.. you are right Yash - there is nothing different from what you offer :)

So if you are saying you have innovated and offered something drastically different from an HSphere Cluster, than I beg to differ.

good for you - but, you would be wrong..

Also, your load-balancing is clearly limited to Linux hosting, which is an HSphere limitiation :)

actually, we are launching Windows solutions shortly and it will be powered by the hsphere control panel.. again, a cloud separates the application layer from the architecture.. the inherent hsphere issues with load balancing Windows are really irrelevant..

To begin with, I have never seen the utility of load-balancers in a classic, shared hosting scenario (whatever the load per server is). In class shared hosting, the point of failure is usually NOT the hardware (and you know this). Coming to my point, with shared hosting (where websites are hosting by different companies on the same server), the points of failure are largely software related. And for software-related issues, an experienced Server administration team, working 24x7 is required to maintain good uptime. Proactive measuring of the websites, on these servers help! That's what we do, and I'm sure you do too.

Wonderful - what happens when one of your shared customers need 6 CPU worth of resources and 8 GB of RAM Yash? don't even bother answering Yash, as we both know the answer to this..


Additionally, you claim you offer a truly scalable solution. The only people that can claim to do so, are Mosso and true-grid providers which are very much larger than Cartika, have alot more man-power and have a huge technological leap over your services. Scalable storage means nothing these days (in a shared hosting environment again) when you can't scale the processing power and RAM on your servers infinitely too!

Yash, are you honestly trying to equate company size into validity of architecture?

So, Mosso is a fraction of the size of Google and Amazon - am I to assume this somehow changes their architecture that they are offering?

Very simply Andrew, your services are not fully scalable and are in fact no more scalable than any HSphere cluster's services.

very simple Yash - incorrect

And i would argue that load-balancing doesn't add any significant uptime advantage :)

please tell me you did not just say this..

Wonder why Amazon offers load balancing with their cloud.. actually, they tell us :)

"Elastic Load Balancing – Elastic Load Balancing automatically distributes incoming application traffic across multiple Amazon EC2 instances. It enables you to achieve even greater fault tolerance in your applications, seamlessly providing the amount of load balancing capacity needed in response to incoming application traffic. Elastic Load Balancing detects unhealthy instances within a pool and automatically reroutes traffic to healthy instances until the unhealthy instances have been restored. You can enable Elastic Load Balancing within a single Availability Zone or across multiple zones for even more consistent application performance. Amazon CloudWatch can be used to capture a specific Elastic Load Balancer’s operational metrics, such as request count and request latency, at no additional cost beyond Elastic Load Balancing fees. See Elastic Load Balancing for more details."

Our HSphere cluster is just as scalable, and we too have consistently delivered great uptime and service at our price point

I do not see you offering cloud based offerings on hsphere - I do not see you integrating Citrix Xenserver into hsphere, I do not see you load balancing services across multiple virtual and physical nodes, I do not see you offering clients the ability to burst beyound the limitations of a single server.. as such Yash - sorry to tell you, but your cluster is not as scalable as our cloud system - and until you invest in the technology and work force, the networks, the software, the hardware, etc to achieve something like this, your cluster will not be as scalable as our HAL offering -

Launching, running, maintaining and evolving this sort of system is a significant investment Yash - and a great challenge - I encourage and challenge you to try - instead of sitting back and trying to minimize our efforts and our offering..

actually Yash - our architecture is pretty close to the MT grid and the Mosso cloud.. funny enough, they also have a control panel on top of their architecture..



Again, you choose Marketing hype to distort the facts and are trying to potray your HSphere-based reseller solution as a grid solution, which it IS NOT!


sure hsphere supports load balancers and SAN's.. last I checked though, hsphere does not support mysql clustering, mysql farms -

MySQL clustering is supported, mysql load-balancing and farms are supported within MySQL!


nor does it support Xenserver and virtualization

This is a different product! and for the record, we too offer a virtualization solutions


"Citrix XenServer™ is the only enterprise-class, cloud-proven server virtualization platform that delivers the critical features of live migration and centralized multi-server management..."

What does this have to do with your Hsphere Reseller HAL Cluster! Your Hsphere Reseller Cluster is not a virtualized, grid platform with live migration! This is a separate product! You are twisting facts to make it look like your HAL hosting is web hosting on a grid! It's still a HSphere cluster though, right?

And for the record, we too offer a scalable, enterprise-class Virtualization platform, with live migration and centralized-server management - its called Virtuozzo. I don't want to get into a Virtuozzo vs XenServer war.. there are plenty of threads here but let's say that both virtualization platforms have its benefits! By the way, we have been doing this for over 3 years now! To me, it would seem like you are a bit late here with the whole Virtualization boom :)



We are actually in the process of integrating this within hsphere so resellers can allocate and sell Xenserver based clouds running hsphere, right from their control panel.. you are right Yash - there is nothing different from what you offer :)

Once again, you are twisting facts! So, you plan to offer VPS accounts (and Hsphere installed on these VPS accounts) serviced from HSphere! This is nothing new! And the fact remains, your HAL Hosting still runs off a a clustered Hsphere platform.. no virtualization.. no grid!

And.. we too offer a Virtualized accounts directly from Hsphere, with similar features, without the marketing hype! Hsphere is a wonderful control panel, and providing what you mentioned is built into Hsphere!


Wonderful - what happens when one of your shared customers need 6 CPU worth of resources and 8 GB of RAM Yash? don't even bother answering Yash, as we both know the answer to this..

We'd setup a dedicated cluster for them! Infact, we have a number of demanding, dedicated server clients who we have satisfied and handled their growing demand!
Since they have a single application, load-balancing and SAN storage would truly make it an effective solution.

Andrew, are you implying that your HSphere HAL Reseller Cluster, could handle a website with 8GB of RAM and 6CPU worth of processing, along with other websites on your standard HAL Cluster?
Once again, your cluster isn't a grid! Stop selling it like that!




Wonder why Amazon offers load balancing with their cloud.. actually, they tell us :)


Once again, Amazon offers a true-grid service! You don't! Load-balancing is intrinsic to the success of a grid! You are applying load-balancing to a shared-hosting model, which has none of those benefits

Yash,

I am not going to go back and forth with you any longer.. I completely disagree with pretty much everything you have said - and I am very very disappointed with the direction you have taken here..

If you want to try and minimize our offerings, wonderful - that is certainly your right to do this..

Try and convince yourself and everyone else you offer the same thing - I actually prefer if you do this - it does not take a genious to figure out you are out on a limb here..

We recently (very recently) changed the name of our offering from HAL to the HAL Cloud - dont you think there is a reason we did that? Keep watching our site Yash - and thank you very much for your feedback and comments -

tcstatic, I am glad you are a loyal Cartika customer :)
Your post is nothing but a personal attack on me, which you are free to do of course.
However, my post is not an attack, rather objective debate.

If you look through my post history you will see I have mentioned/recommended your company in the past when it comes to H-Sphere or clustered server environments, so paint me however you will. It is not an attack, it's simply saying you had absolutely no business interjecting yourself in this thread in the first place with deceiving comments soliciting your services in post#4. Then go on to say you offer the exact same thing or something similar when you don't. When you lose that, you then go on to say there is no value in a load balanced fully redundant hardware/services solution. I see you also do not see the value in several other items/services that are unavailable in your offerings..this is your right and is the choice of the customer. But at least deal with facts.

I would think as a business owner you would have learned a long time ago that if you have to try to tear down a company to make yourself look better ....you've already lost.


Comment/Rant: I sure miss the days when WHT was a place where mutual respect was given by most hosts to each other. Of course those were the days before a poster was preyed upon by vultures using spam sigs, and being told they "need" a VPS for a simple PHP/MySQL site.

Andrew, I believe you seek to cause confusion by relating your "HAL" services with a true grid platform like Mosso.

Your HAL service is a simple extension of HSphere. Hsphere already has the ability for SAN storage (which is nothing new in the industry) and load-balancers. So if you are saying you have innovated and offered something drastically different from an HSphere Cluster, than I beg to differ. Also, your load-balancing is clearly limited to Linux hosting, which is an HSphere limitiation :)

I am very well educated with the terms load-balancing, cloud networks and their capabilities. I have worked on a very large corporate cloud network as a Software Engineer (at Cisco Systems) and have seen load-balancers in action!

To begin with, I have never seen the utility of load-balancers in a classic, shared hosting scenario (whatever the load per server is). In class shared hosting, the point of failure is usually NOT the hardware (and you know this). We run quality, redundant hardware at JodoHost (such as redundant disk arrays, servers with redundant parts and a network with redundant routers, etc). I am sure you run a similar, well maintain hardware/metwork :) Coming to my point, with shared hosting (where websites are hosting by different companies on the same server), the points of failure are largely software related. And for software-related issues, an experienced Server administration team, working 24x7 is required to maintain good uptime. Proactive measuring of the websites, on these servers help! That's what we do, and I'm sure you do too.

So I'd claim, that load-balanced web servers (which you dont offer on Windows) will not improve uptime significantly. We've maintained 99.9%+ uptime on our servers, with simply redundant hardware and a redundant network. Load-balancing best works when a SINGLE application has to be scaled over a several servers.. Load-balancing was never really designed/intended for several or 50-100 independent domains being load-balanced on a server, as its simply not as effective!

Additionally, you claim you offer a truly scalable solution. The only people that can claim to do so, are Mosso and true-grid providers which are very much larger than Cartika, have alot more man-power and have a huge technological leap over your services. Scalable storage means nothing these days (in a shared hosting environment again) when you can't scale the processing power and RAM on your servers infinitely too!

Very simply Andrew, your services are not fully scalable and are in fact no more scalable than any HSphere cluster's services. And i would argue that load-balancing doesn't add any significant uptime advantage :) Our HSphere cluster is just as scalable, and we too have consistently delivered great uptime and service at our price point

That was VERY childish!

Cartika's HAL is exactly like Mosso's system: SHARED CLOUD HOSTING - which is scalable and offers 99.99% uptime instead of 99.9% uptime.

Why a site in Mosso can scale and in Cartika cannot scale? If more resources are needed both providers plug the servers and you are ready to scale!

Your Hsphere cluster is just as scalable? Ok, in 1 hour a website hosted on your h-sphere cluster (or in Cartika's standard cluster) will accept 100k visitors. The website is limited on 1 webserver so it cannot accept the load and all the connections will time out. In Mosso or in Cartika's HAL the website will be still functional as they will plug more servers in that array of the webservers.

01globalnet,
Mosso = virtualized grid platform
Cartika HAL Reseller = HSphere Cluster with load-balancing on the web-nodes.

Mosso can scale dynamically.. A load-balanced HSphere cluster cannot.. At best, you need to add servers manually which is far from a dynamically scalable solution and will still suffer from downtime incidents if a single website had the requirements of an entire server.. (software problems, which load-balancers don't check for!). Remember, this single large website (with the requirements of a server) is still sharing physical resources with the other websites... that's not like a grid!

Load-balancing in a shared-hosting environment will not produce the same results as on a virtualized grid, that was load-balanced. In a virtualized grid, each applicataion is isolated from software failures that may result from the sharing of resources with its neighbors. In a shared-hosting environment, they are not. A virtualized account can take over the entire server.. and other accounts can be moved off easily and even on-demand/dynamically.. in an HSphere cluster (like Cartika's) they CANNOT!

There is a huge difference there. However, I am no longer going to continue this discussion as the technical differences are of course not clear to those that get taken by marketing hype

Wow, I'm not going to take sides here but as a person who has actually tested/used Cartika's HAL solution I found it to be quite a good solution for anyone who needs a high availability solution at the fraction of the cost of setting up your own cluster.

If you've followed Mosso's solution much you have seen a number of bumps in the road as they work it out and people have had some bad experiences and even moved off it. They are working out their kinks and will have a good solution but it is basically the same technologies as any other clustered host: load balancer, server, san, etc.

To say the two solutions aren't alike seems kind of silly to me and to say Mosso is a "grid" solution seems like someone needs to look up the definition of grid computing.

Now go back to fighting amongst yourselves, thanks!

Wow, I'm not going to take sides here but as a person who has actually tested/used Cartika's HAL solution I found it to be quite a good solution for anyone who needs a high availability solution at the fraction of the cost of setting up your own cluster.

If you've followed Mosso's solution much you have seen a number of bumps in the road as they work it out and people have had some bad experiences and even moved off it. They are working out their kinks and will have a good solution but it is basically the same technologies as any other clustered host: load balancer, server, san, etc.

To say the two solutions aren't alike seems kind of silly to me and to say Mosso is a "grid" solution seems like someone needs to look up the definition of grid computing.


Thanks for the sanity check Russ - glad to hear someone with actual experience piping in..

However, I am no longer going to continue this discussion as the technical differences are of course not clear to those that get taken by marketing hype

I am actually somewhat shocked at what has gone on here.. Yash - you and I have always maintained a professional and "friendly" competitive relationship.. How you suddenly regressed to the point of jumping into a thread you do not belong in and attacking us - is simply beyound me..

Having said this Yash, if you want to understand our technology, you only need ask.. maybe this would have been smarter then just throwing around accussations and insults - you have made yourself (and unfortunately me by default) - look like total @$$'s - it is obvious that your understanding of our solution is limited to your beliefs of what hsphere is capable of (probably what your parallels support person has told you).. Eitherway, hsphere is just a control panel - it is merely a GUI that enables users to create accounts, add domains, etc.. it in no way limits or impacts the architecture it is run on - well, at least if done properly.. we are able to increase or decrease sizes of containers, we are able to integrate our Citrix XENserver cloud layer with our Citrix net app load balancers and create cloud systems with our without load balancing - we are able to deploy additional virtual or physical nodes within minutes, etc, etc, etc.. again, just because you do not understand this, or you do not know how to do this, does not mean it is not possible - it simply means you dont understand..

I have watched this industry develop for almost 2 decades - I have watched Mediatemple launch a grid and bring their first version end of life and launch a new solution on new architecture because of various problems with their solution, I have seen hosting solutions develop from servers with admins manually configuring everything to automated solutions spreading accounts over fleets of servers, replicating data, load balancing requests, organically adding resources to that users sites can grow and seamlessly expand beyound the physical limitations of a server, etc..

Our HAL system and our HAL cloud is an evolving technology - it is leaps and bounds further then what you Yash are giving us credit for and it is leaps and bounds away from where I want it to end up...

My only message to you Yash, would be that you should focus your efforts and your energies on being competitive and stepping up with equal or better solutions - we have well over $250k and countless man hours and integration time invested in our Cloud offering - the investment is just going to increase - sitting back and baselessly and flagrantly assaulting us will not suddenly make you competitive Yash - step up or shut up...

Yash,

technology does not matter here. BOTH Mosso and Cartika offer Cloud or Grid or Load Balanced or Whatever hosting. Each provider is using different hardware and software to achieve this.

And both providers sell SHARED cloud hosting (as I know Mosso does not isolate each website in a virtual container) - so a shared cloud environment has the same limitations and problems as any shared account.

And both providers sell SHARED cloud hosting (as I know Mosso does not isolate each website in a virtual container) - so a shared cloud environment has the same limitations and problems as any shared account.

I wouldn't say that at all. With a cloud/grid/loud-balanced shared hosting platform the idea is that your website can scale without the need for you, the site owner to worry about managing those load-balancers and clusters of servers. Most end users don't want to worry about scaling their application, and deploying and configuring virtual servers (e.g. EC2). You just want to install your wordpress blog as you would do on any system and get on with making your website a success, in the knowledge that if you are Dugg or Slashdotted, your site will stay up.

This is not the case with standard shared hosting environments.

This new movement is attempting to be a best of both worlds. The ease of use of shared hosting for the end-user, with the reliability and scalability and other features that come with a cloud/grid/high availability platform.

I do not disagree :) actually I agree :)

~ the same limitations ~ I mean as opposed to Yash's post shared vs dedicated/virtual and problems caused because of 'bad' neighbours.

we use both jodo and cartika
cartika's support is FAR FAR faster
at jodo, it OFTEN takes hours to get a reply