| Posted by developer_designer, 04-02-2017, 02:09 PM | |
Hello,
I moved from shared hosting to a VPS over a year ago. I went from $14 a month shared to $67 for a vps with litespeed. Recently WiredTree ,my VPS host was acquired by LiquidWeb which made me take a look at my hosting again. After some testing I found a $4 a month shared account at HawkHost that tested up to two times faster than my VPS through pingdom tools and dotcommonitor. I have seen some better results worldwide by using Cloudflare as well.
I have read some posts here that a good shared host can be better than a vps because you may be utilizing 24 processors instead of 1 or 2 if your shared environment isn't oversold.
The problems with shared hosting seem to be bad neighbors who overuse, or allow security breaches.
Does anyone have any comments on this? If you don't need access to the root for customization, meaning you don't need a vps to install custom software, what do you have to pay to move up from shared hosting and get better speeds.
Is there unseen throttling that occurs in shared hosting? If you have 300 simultaneous users hitting your site, does the server slow your site down somehow? or is the metric for that way high?
Thanks
|
| Posted by TrentaHost, 04-02-2017, 04:43 PM | |
Every hosts Shared Hosting will differ, some oversell some don't, some optimize better then others in terms of usage. a VPS would definitely be faster then a shared host in the amount of resources allocated, but you need to keep in consideration that a VPS need's to be optimized, perhaps yours wasn't as optimized as a Shared Hosted would optimize their servers.
|
| Posted by JSCL, 04-02-2017, 05:11 PM | |
The perfect balance between a VPS and Shared Hosting is Elastic Sites. You should look in to it and the benefits it gives you - https://elasticsites.com/
Dedicated resources like a VPS, but in a Shared Hosting environment. There are a number of providers offering Elastic Sites too.
|
| Posted by WHE Tom, 04-02-2017, 06:24 PM | |
I guess the key is if the shared hosting envrioment is oversold or not. If you take into account that a shared hosting envrioment is on the verge of being at a max capacity then your VPS will definately be much better in terms of performance. Some hosts may throttle users if their usage is much higher than average, but it really depends on the individual company etc.
|
| Posted by BoomHost-Kumar, 04-02-2017, 09:39 PM | |
The above statement is a general assumption when people think of shared hosting but things have changed in shared hosting from those Apache days to better or more controlled environment like CloudLinux OS. Whatever the case is it all comes to how closely the servers (and network) are monitored and administered by the hosting companies.
|
| Posted by MechanicWeb-shoss, 04-03-2017, 02:17 AM | |
A typical shared hosting can be faster in many cases when you are hosting multiple websites. Many shared hosting providers offer cPanel on a CloudLinux based system where CloudLinux offers light-weight isolation for each hosting account on the server. On the other hand, resources (CPU/RAM/IO) are allocated on a per cPanel basis. Generally speaking, if you are hosting 10 websites on a shared hosting, each under its own separate account and each account has 1 GB RAM, you will have a total of 10 GB RAM at your expense. If you put all 10 websites on a 4 GB VPS, you have far less RAM per website and performance could degrade.
But if you have one or two resource hungry websites, a VPS should be the way to go. Hope this clarifies.
|
| Posted by gagah, 04-03-2017, 05:10 AM | |
If you are using pingdom tools to benchmark your site's performance, you should be able to see where the bottleneck is with the waterfall tools. What kind of resources load the longest? Where is your bottleneck? Two times faster is a big performance difference, and indicates more of a lack of optimization rather than the types of hosting being used to me.
But to some extent, yes, shared hosting can work better, since if they are not oversold, they do have decent ability to burst. Like others already mentioned, typical cPanel shared host node will have much more ram and cores than a low-specced VPS, but this is such a subjective "your mileage may vary" kinds of experience and not really a definite thing to hold on to, since in a shared node, load can varies greatly from time to time without any of your personal input, while in a VPS, you control the whole environment yourself (if it's KVM), and performance is more predictable and linear in the right hands.
|
| Posted by kpmedia, 04-03-2017, 06:36 AM | |
Shared can be faster than VPS, yes. It depends on several factors.
More important = the quality of the host.
|
| Posted by LJSHost, 04-03-2017, 07:00 AM | |
As others have said, it depends shared hosting can be faster but this all depends on the provider. I think it all comes down to overselling and keeping a balance.
|
| Posted by developer_designer, 04-03-2017, 12:11 PM | |
Thanks for all the feedback,
With all the possible variables it seems like unless someone has experience in server admin, the only way to know is live testing. I went with a recommended managed VPS host added litespeed, asked that it be configured to deliver php as fast as possible (not sure what that might be). My site is just simple php includes, no database and yet it really wasn't super fast, there was nothing bad about it, i just didn't expect to beat it so easily with shared. For the most the uptime was excellent.
As a hosting consumer its makes the choices difficult to value.
Are pingdom tools and dotcommonitor the best tools to check? pings just check server response but not rendering and delivering the whole page.
|
| Posted by LJSHost, 04-03-2017, 12:17 PM | |
That does not sound good, a simple site with no database intensive usage should have a minimal impact on the server.
Even a 1 core 500MB RAM VPS should deal with a lot of traffic without issue on that sort of site. I suspect you have a underlying issue but could not comment more without seeing your top output.
I think https://gtmetrix.com/ is more what you are looking for perhaps.
|
| Posted by mjfleming, 04-03-2017, 03:01 PM | |
Application level inefficiency goes way beyond database usage, wouldn't you agree?
When you say your website is "all php includes", you mean it is just php includes? That's going to be basically just a lot of disk i/o, if you are being hosted on ssd you should be fine?
A virtual private server would only be faster if you know how to tune it properly! But also I think you are comparing a zero host to a pretty darned good one (hh get good reviews around here I noticed)
|
| Posted by gagah, 04-03-2017, 03:07 PM | |
Looks like the VPS provider that OP is comparing HH to is Liquidweb, not exactly a zero host, and it's managed too.
|
| Posted by Julien@Hostabulous, 04-03-2017, 03:15 PM | |
So basically anyone running cloudlinux.
|
| Posted by mjfleming, 04-03-2017, 04:00 PM | |
Well, to be sure they have alot of negative stuff floating around right now after migrating WT customers right?
|
| Posted by gagah, 04-03-2017, 04:06 PM | |
Indeed, but my point is they're still a pretty reputable company, so having similar level of expectations to an unknown host isn't exactly right, and even on price point alone i'd consider LW a mid to high end shop, and should be expected to perform at least to the same level as a company like HH, especially against their shared products. I might be wrong though.
|
| Posted by mjfleming, 04-03-2017, 04:10 PM | |
Fair enough, I do know they are supposed to be mid-to-high, I agree there. But just based on the torrent of stuff I've seen on the thread here and all over reddit about LW I'm inclined to believe this may be more of the same.
|
| Posted by TierNet-Nate, 04-03-2017, 04:41 PM | |
As many people have already said, much of the speed and performance come down to two things. The quality of the host providing the service and the optimization of the site and service. If you get a poor host with a poorly managed service the performance will be bad. If you get a decent host with poor service management then again your performance will be bad. If you get a good host with good service optimization and management you will see good performance. I'd say check reviews here and see if you can find a good host that takes care of their customers and provides good management for the service so that you can get the most optimized performance.
|
| Posted by developer_designer, 04-03-2017, 05:01 PM | |
Yes I had WT. I had researched them at the time and by all accounts they were worth the cost. I only looked around because I felt I needed to research LW because of the move or start again. In doing testing with a migrated account I noticed better performance on HH shared.
|
| Posted by KontikiWebDesign, 04-03-2017, 06:56 PM | |
WT was worth the price. We have some customer not complaining at all moving from the WT to the LiquidWeb family... And if you ask them they say that Most of WT VPS options used shared CPU. Their Storm VPS offers dedicated CPU.
So probably you should try to optimize your VPS or ask them before you move again...
As direct experience i can also recommend you MightWeb Hosting and you can ask about Marcus. He can handle the kind of question you made about his hosting and you can check with him about the best solution that better fits your needs
|
| Posted by DewlanceHosting, 04-04-2017, 12:29 AM | |
In terms of stability, VPS will be much better but speed depend on upstream provider. For example If your shared hosting server is nearby pingdom testing server then of-course speed will show much better than your VPS which is located to far from pingdom speed test server.
However, VPS speed can be slow due to overselling, overload, etc., factors.
Try to optimize your VPS speed, use less images or compressed images on your site.
|
| Posted by kpmedia, 04-04-2017, 01:25 AM | |
... to not know at all.
Simpleton "tests" are often full of variables that skew results.
How big was the VPS? Did it have cPanel? That alone can determine speed of the VPS. But even more important is the hardware of the VPS, or the virtualization method (example: Xen vs OpenVZ), versus the specs of the dedicated (or cloud VPS) used for the shared.
I'm bet more on CPU or RAM.
I'm betting the VPS is not cached or optimized at all (customer at fault, not host), while their own shared servers would be highly optimized. If the VPS is managed, you need to ask for help.
|
| Posted by mjfleming, 04-04-2017, 01:29 AM | |
php's include function is not hard on ram or cpu especially since he is on a vm which is "managed", if it has a sane config it wouldn't even be compiling those scripts once an hour. even with management that vps has to have at least enough ram to make opcode caching a non-issue
the original poster doesn't need to move necessarily, his vps is probably lightyears ahead of most shared hosting for his usecase, but it is preferable to move back to shared hosting if the scope of your application is currently just php includes-- this is for all intent and purpose a static site. Nothing wrong with that, but it's not going to hog resources unless you're doing really dopey things with your php.
Last edited by mjfleming; 04-04-2017 at 01:33 AM.
|
| Posted by kpmedia, 04-04-2017, 01:38 AM | |
I'm looking deeper than you -- the actual php that's being included.
|
| Posted by AdelinaHost, 04-04-2017, 03:25 AM | |
I can say that shared hosting is faster than VPS with 256/512 RAM, however VPS is needed if you have some special settings and modules which you cannot have in shared hosting.
|
| Posted by Host4Geeks-Kushal, 04-04-2017, 04:18 AM | |
As rightly said by many, just moving to a VPS will not guarantee magical improvement in performance or site load times. A well managed server with a good host with can and will offer superior performance when comparing with a VPS on a oversold / poorly managed node. The question to ask here is, does your website need a VPS? What is your resource usage in terms of CPU, RAM, IO? Is your site optimized properly? Run some GTMetrix and Pingdom tests and see how can you shave off some ms off the load times.
|
| Posted by IH-Jake, 04-04-2017, 06:21 AM | |
Yes, it is possible. It depends on the core server. If the VPS is hosted on an overloaded node then it can affect the performance. At the same time, if the shared server is under loaded and well maintained it could give you good performance.
To keep your domains isolated from neighbors, you can use a dedicated IP to use as the mail server IP for your domain as well. A server utilizing cloudlinux can give you assurance about the resource usage as well. If you have more than one domain, go for a reseller account.
|
| Posted by x-flow, 04-05-2017, 10:15 AM | |
Generally speaking, a VPS come with better performance than shared hosting since the VPS has much higher aligned resource.
However, if a physical server hosts too much VPS, or VPS has very high CPU/RAM/IO usage, the performance could fall, in some cases, you may see a VPS is slower than shared hosting.
|
| Posted by singumgum16, 05-03-2017, 11:09 PM | |
I think vps faster hosting..
|
|
 Add to Favourites
 Print this Article
|
