Print Page | Close Window

Data Center Problem

Printed From: RomanCart.com Forum
Category: RomanCart
Forum Name: Announcements
Forum Description: Important RomanCart News
URL: http://forums.romancart.com/forum_posts.asp?TID=13520
Printed Date: 19 April 2018 at 11:48pm


Topic: Data Center Problem
Posted By: Support
Subject: Data Center Problem
Date Posted: 10 September 2016 at 11:28am
Hi,

The data center we use has a problem affecting our services.

We are waiting for them to clarify and resolve the issue.

Apologies for the obvious inconvenience.

Support



Replies:
Posted By: Support
Date Posted: 10 September 2016 at 12:24pm
Hi,

It has been confirmed that it is a power issue at the datacentre.
After the last power issue redundancy was added and it seemed very unlikely this could happen again, however unfortunately it is out of our control.

We are waiting for power to be restored.

Please follow us on twitter @romancart where we will provide updates.

Apologies once again,

Support


Posted By: FuzzyOne
Date Posted: 10 September 2016 at 12:47pm
What DC are you in so we can get updates direct from them?

How in the hell do you add redundancy but rely on the same DC?

What are you going to do to rectify this situation in future?

It's pretty evident to work out what your income is via your linear invoice numbers, are you investing into this system or running it into the ground, you had a good thing going but it seems like little investment into any of your systems has allowed you to go stale, and while this downtime might be fine for low volume customers it's beyond forgiveable for your high volume customers.


Posted By: Support
Date Posted: 10 September 2016 at 1:12pm
Hi,

We have been told that power should be restored in the next 30 minutes.

FuzzyOne - we have invested in a completely new hardware stack which we migrated to recently.

Unfortunately our business model could not sustain a multi data center operation. (Which brings its own problems.)

The datacentre has multiple diesel generators, a flywheel UPS and other UPS's which they added to as a result of the previous problems.

We have to rely on the datacentre to resolve these issues.
We could change datacenters however one would hope that a DC which has suffered 2 failures would surely work out what has gone wrong.

Apologies for the inconvenience,

Support




Posted By: rscass
Date Posted: 10 September 2016 at 1:13pm
Global Switch. Their website isn't working either. Looks like they didn't think to host their customer support site at another location.

Not much point in having a cloud based system if you attach all of your 'cloud' servers to the same bloody plug socket (even if you did put a bigger fuse in it in June). I do hope Romancart will be asking them for a refund.


Posted By: rscass
Date Posted: 10 September 2016 at 1:16pm
Hopefully they'll invest in a bigger hamster and wheel.


Posted By: softap
Date Posted: 10 September 2016 at 1:18pm
Working OK now - 13:18 GMT


Posted By: Support
Date Posted: 10 September 2016 at 1:19pm
Hi,
Please be prepared for slow performance for the first few minutes whilst lots of people try and log in and the system catches up.

Many Thanks
Support


Posted By: RowanEng
Date Posted: 10 September 2016 at 2:07pm
Support,

To say you "hope the DC will work out what went wrong" just isn't an acceptable answer.

Romancart customers (that's us here on this forum) need to know what you are doing about these problems.
What plans will be put in place?
Will you give a solid assurance outages like this won't happen again in the future?


Dave.


Posted By: Support
Date Posted: 10 September 2016 at 2:15pm
Hi,

We have to rely on a datacentre for power and connectivity.
Our chosen datacentre is Claranet at Global Switch 2 which has given great service for around 10 years.

The questions you are asking are questions that we will be asking Claranet/Global Switch 2.

Unfortunately they and us can never give guarantees regarding this type of problem. In fact, the best you ever get is 99.99% uptime which still allows for significant downtime.

Ultimately, of course, if this kept happening we would move to a different datacentre, however one could argue that a datacentre which has had and resolved issues may be more resilient than a datacentre which has never had problems. Moving to a different datacentre brings issues of its own.

Many Thanks
Support



Posted By: EddieODG
Date Posted: 10 September 2016 at 2:43pm
It has been disappointing this morning, and a Saturday makes it worse for many in eCommerce. However as stated, a 99.9% uptime which is pretty standard still leaves a lot of potential downtime. Normally problems are quickly resolved, however this appears to have been a larger problem, and I am sure the datacentres will put in place contingencies, should a similar problem occur.

It is unfortunate, but one of those things as far as I can see.

Pleased it is now resolved.


Posted By: rscass
Date Posted: 10 September 2016 at 10:47pm
Good news. Global Switch have managed to get their own website back online.

According to their homepage, their tag line is "Global Switch, Always On".

Worth a read.

"Global Switch data centres operate to a Tier III standard or higher. The power infrastructure is designed to ensure robust resilience with "no break" IT power. We constantly invest in power capacity and adopt a policy of pass-through electricity costs based upon consumption only"

No break? Really? I'd love to hear them selling this concept to Claranet/Romancart on Monday morning.

99.99% uptime allows for 52.96 minutes of down time in any 365.25 day period. On this year's performance, I reckon they need to guarantee 100% uptime for approximately seven years in order to put things right.

Just saying...



Print Page | Close Window