How are you finding eNlight?

[Ben Stones]

Hi everyone,

We'd like to thank all customers for their feedback so far on our new flagship cloud hosting platform. We're really excited about the future of eNlight. It's an innovative solution, and kudos to our eNlight engineers and other employees involved in the development of our cloud hosting platform.

That aside, I'd definitely be interested to hear what you think of your eNlight Cloud so far, and anything you would like to suggest to our eNlight engineers. How are you finding eNlight? Performance, cost, and so forth. We appreciate feedback from our customers.

We have had some important suggestions from some of our customers already, and have or will be implementing further improvements and features to eNlight. One of the improvements we made pretty quickly was regarding "performance modes", allowing customers to choose between a performance mode and economy mode with their eNlight VM(s). Economy mode uses as little resources as possible; clears cache and other unnecessary data which can ultimately result in even lower costs for your VM use. And customers can choose performance mode if they wish to have the best performance with no compromises. As eNlight works based on what you use, and personally I think the great thing about eNlight is that it is the same sort of system Pay As You Go mobile SIMs work, where you top up when you like, with whatever amount you like. And if you run out of credit, your service simply stops.

So as an eNlight Cloud customer, how are you finding it?

[triton1]

Fast, smooth, and an excellent platform for my needs, My opinion you just need some how to's posted here, on setting up the VM part of it.

[ejsolutions]

How are you finding eNlight?

Well, I went to the Home page of eUKHost and lo & behold...

I still find that the initial response to a HTTP request is slow - like it needs wakened up (UKR3). Unfortunately, other commitments haven't allowed me time to migrate active sites over to it, yet - I will get there in the end. Hopefully, I'll have a better idea of performance then. I do use one site on eNlight sporadically, for osCmax button creation, and once the connection is established, everything is fine.
Response to logging into WHM/cPanel is slow compared to other Clouds.

EJ

[P.S. Just been checking WHM on a number of Cloud instances and I'm pleased to report that response appears to have been greatly improved! ]

[ejsolutions]

...you just need some how to's posted here, on setting up the VM part of it.

The initial VM setup is very intuitive, IMO, unless you mean specifically when a cPanel environment isn't chosen and then the configuration of the OS.

[triton1]

The initial VM setup is very intuitive, IMO, unless you mean specifically when a cPanel environment isn't chosen and then the configuration of the OS.

where i got lost is when WHM was updated it changed my password I was always under the assumption that you had to change it in whm but found out it can be changed in enlight panel, handy feature but i'm old so sometimes these new fanged things confuse me

[ebusiness]

I have an eNlight account - I was lucky enough to find an 'Easter Egg' so it persuaded me to try the service. My business has had an Eukhost account for 5 years or thereabout and I have hosting accounts and a VPS with other hosting companies..

The concept of pay as you go web services with the ability to scale on demand up to the pre defined limits is excellent and a valuable service for an online business.

I have some observations of the service so far - please take them as feedback not a criticism - I am very happy with the service so far.

The 'predict your costs' tool was confusing. It states "You will be charged only for your base specification and then for the additional resources you use. " I debated what base specification I should select whereas it is not relevant. My experience was pay a £15 set up fee and thereafter there is no cost until resources are used. Create a VM and set the cpu and ram parameters as upper limits - this caps your maximum cost.

Some guides on VM / WHM set up would be very helpful and avoid chat sessions with support. After opening the account I tried to set up a VM and when logging into WHM learned that the 2 nameservers pre coded in the template seem to be dummies. Support then confirmed that I needed to supply my own private nameservers. This took a while to set up due to propagation delays and is not mentioned in the pre sales comments.

I was going to be critical that the template did not have csf / lfd installed and running (security) and other elements were out of date - php - I set about installing / upgrading these myself. I then read your note of today and realise that this is a fully managed service - I had not picked that up in the sales material and thought that this was a service where I was on my own rather than , as is the case, I had access to a support team.

I am running a 64bit Centos Cpanel template with 1024 ram and 4 cpu. It has a couple of low activity web sites at the moment. As minimum hard disk is set at 20gb the daily running costs are 44p excluding the Cpanel license. A 32 bit Centos Cpanel template would be beneficial ( lower ram usage) and a 15 gb or less hard disk option would make this a very keenly priced service ( but I recognise you need to make a profit!)

HTTP response is on a par with other platforms I use but this is surprising as this one has more power available. Would like to understand how to tune this platform or receive some insight as to how it will operate when more heavily loaded. If I brought across some more heavily used sites and increased the ram / cpu caps would the machine start to perform better?

Reading the forum and noting the engagement of the management / development team convinced me to get involved with eNlight - it is good to see the team so focused on making this a success. (It also encouraged me to pen this note).

PS - To be confident running this as a business operation I need to be able to backup the VM / snapshot the VM before changes and practice recovery. When will these functions be available?

[ghosthosting]

I love the whole concept of it - have just built a Windows server as an Exchange server and it was a bit of a nightmare, but finally up and running.

Am definitely interested in the API though - I'd love to resell the whole concept of PAYG hosting and will be looking forward to seeing how we can integrate enlight into WHMCS.

[Ben Stones]

I love the whole concept of it - have just built a Windows server as an Exchange server and it was a bit of a nightmare, but finally up and running.

Am definitely interested in the API though - I'd love to resell the whole concept of PAYG hosting and will be looking forward to seeing how we can integrate enlight into WHMCS.

Hi,

I will get an update from Rishi on the status of our in-development eNlight API and get back to you.

Thanks for your feedback on eNlight.

Kind regards,
Ben.

[eUK-Rishi]

I love the whole concept of it - have just built a Windows server as an Exchange server and it was a bit of a nightmare, but finally up and running.

Am definitely interested in the API though - I'd love to resell the whole concept of PAYG hosting and will be looking forward to seeing how we can integrate enlight into WHMCS.

Thanks for the Feedback

We are in process to develop the Public APIs for eNlight Cloud and should be live in two months. A WHMCS plugin is also proposed for eNlight and our Billing Dev team is working on the possibilities and feature scope for it.

Stat tuned for further updates ..

[Emmelle]

My eNlight is currently completely offline as per ticket IQJ-924-59776. No update on Live Chat, no reply to ticket, nothing on euk-status.com Scarily silent from eUK considering my VM has just disappeared.

[eUK-Martin]

Dear Emmlelle,

We are facing few problems with UKR3 and we are working on it to make cloud completely operational. The status pages been updated and more details will come along it as soon as we hear from our technical team.

Thank you for being patient, it is appreciated.

[ejsolutions]

My eNlight is currently completely offline as per ticket IQJ-924-59776. No update on Live Chat, no reply to ticket, nothing on euk-status.com Scarily silent from eUK considering my VM has just disappeared.

I noticed spurious I/O wait issues and Load appreciably higher than normal, yesterday evening but got sidetracked into not reporting it. Glad to see that the node appears to have been 'fixed'.

[Side note: Swap space usage (RAM allocation) is still greater than I would expect on a Centos 512Mb RAM Virtual Machine. If Rishi has time perhaps he could look at my Munin charts - no major rush, though.]

[Emmelle]

The status page was only updated after I asked for it to be done so on Live Chat. What's the point of having a status page if it isn't relevant until someone asks for it to be updated. Also, when will we be getting an explanation of why this multi-homed, 100% uptime service was down for so long? If it was storage, why wasn't a redundant copy brought in, if it was the hosts, why didn't they auto migrate to new hosts etc.

Though this is supposed to be one of the most resilient platforms, I am still feeling to need to have a live redundant VM either on another node, or even with another host. This is far from ideal.

[John]

Hello,

This was an issue with the master node of UKR3. The cause is still being investigated and further updates will be posted when more is known. The lack of updates will also be addressed. It wasn't emergency maintenance.

[John]

I noticed spurious I/O wait issues and Load appreciably higher than normal, yesterday evening but got sidetracked into not reporting it. Glad to see that the node appears to have been 'fixed'.

[Side note: Swap space usage (RAM allocation) is still greater than I would expect on a Centos 512Mb RAM Virtual Machine. If Rishi has time perhaps he could look at my Munin charts - no major rush, though.]

You really do notice everything, don't you There was some prep for the Saturday work being carried out which involves removing some drives and replacing them with SSD in the storage cluster. This started at 17:10 but some slowness was noticed at approximately 19:30 (wait time higher than acceptable) so the task was given a lower priority (takes longer to complete, but less strenuous) at approximately 8:15pm, which rectified the issue.

Unrelated to the actual problem which caused the downtime in the early hours.