CloudatCost and Fibernetics April to May Outage
-
@dafyre said:
They Seriously need to be using Zabbix or something for monitoring their links...
But then they couldn't always blame outages on someone else.
-
Hey everybody,
In the past 24 hours we have experienced multiple fiber splices which have caused a lot of downtime at our data center.
As many of you have mentioned, we are not completely offline and there are servers still online during this outage.
The most recent update I received from Rogers outlines that techs have been dispatched and that the issue has been escalated to management at Rogers.
The crew is currently trying to resolve the issue but Rogers has not provided an ETR.
I will do my best to keep everybody posted on a regular basis.
-
@AmanBhogal said:
Hey everybody,
In the past 24 hours we have experienced multiple fiber splices which have caused a lot of downtime at our data center.
As many of you have mentioned, we are not completely offline and there are servers still online during this outage.
The most recent update I received from Rogers outlines that techs have been dispatched and that the issue has been escalated to management at Rogers.
The crew is currently trying to resolve the issue but Rogers has not provided an ETR.
I will do my best to keep everybody posted on a regular basis.
What does Rogers have to do with it? You have redundant connections, why have you chosen not to failover? The issue is with you, not with Rogers. That there is a loss of redundancy, a loss of capacity, that might be Rogers. But the downtime issue is not Rogers and mentioning them is an excuse.
Rogers has nothing to do with the downtime. Why are you DOWN is the question and why is no attempt at resolving it being made?
-
@Reid-Cooper said:
@AmanBhogal said:
Hey everybody,
In the past 24 hours we have experienced multiple fiber splices which have caused a lot of downtime at our data center.
As many of you have mentioned, we are not completely offline and there are servers still online during this outage.
The most recent update I received from Rogers outlines that techs have been dispatched and that the issue has been escalated to management at Rogers.
The crew is currently trying to resolve the issue but Rogers has not provided an ETR.
I will do my best to keep everybody posted on a regular basis.
What does Rogers have to do with it? You have redundant connections, why have you chosen not to failover? The issue is with you, not with Rogers. That there is a loss of redundancy, a loss of capacity, that might be Rogers. But the downtime issue is not Rogers and mentioning them is an excuse.
Rogers has nothing to do with the downtime. Why are you DOWN is the question and why is no attempt at resolving it being made?
Exactly. How does a Datacenter not know how to advertise it's routes and subnets over multiple connections? people run more reliable and redundant hosting out of their houses.
-
-
-
Back up here.
-
Up here. going to get some stuff off of it. then never use it again. Still waiting on the paypal disputes.
-
Finally back!
-
I'm back online as well.
I won't hold this against them... I have seen and been part of (both the cause and on the repair team) network issues like this (although at a much smaller scale). I just signed up about halfway through April for one of the Dev 2 instances, I think.
Stuff does happen, but if they have failover links, then everything should fail over, or they should make it apparent in the feature list on the pricing page that the Dev & Big Dog instances are not part of the failover links if that is the way they want to go.
-
This is just my 2 cents but I have been using Vultr from beta and they are great.
-
Here is my IOWait from one of the boxes that just got brought back up...
01:53:57 PM all 12.47 0.00 9.12 26.66 0.00 51.75 01:55:07 PM all 23.65 0.00 14.91 37.25 0.00 24.2026% and 37%, those are crazy numbers!
-
Im just going to leave this right here
"2) Service Interruption Credits
If Services are interrupted for a period of at least 24 hours after notice by the Customer to CloudatCost, an allowance equal to 1/30th of any fixed billing cycle charges for the affected Service shall apply to each full 24 hour period during which the interruption continues. Credit in any billing period shall not exceed the total average monthly charges for that period for the affected Service. No allowance shall apply to any non-recurring or usage charges. The Customer's sole and exclusive remedy for Service interruption shall be the credit outlined above
The Customer must request credit within thirty (30) days of the validated interruption, failing which the Customer forever waives its right to any credit for such outage. CloudatCost reserves the right to determine in its sole and absolute discretion if an interruption is considered a validated interruption.
At no time will multiple remedies be provided to the Customer for same, similar or related troubles for the same service. Customer payables must be current prior to the Customer receiving any credits
No credit shall be provided for interruptions due to the following:
the failure of any component, software, network or system provided by the Customer or a third party;
network maintenance by CloudatCost;
Force Majeure events;
acts or omissions, negligent or otherwise, of the Customer or a third party, or any breach of any term or condition by the Customer of the Master Services Agreement or any schedules or addenda attached thereto." -
"2) Service Interruption Credits
If Services are interrupted for a period of at least 24 hours after notice by the Customer to CloudatCost, an allowance equal to 1/30th of any fixed billing cycle charges for the affected Service shall apply to each full 24 hour period during which the interruption continues. Credit in any billing period shall not exceed the total average monthly charges for that period for the affected Service. No allowance shall apply to any non-recurring or usage charges. The Customer's sole and exclusive remedy for Service interruption shall be the credit outlined above
The Customer must request credit within thirty (30) days of the validated interruption, failing which the Customer forever waives its right to any credit for such outage. CloudatCost reserves the right to determine in its sole and absolute discretion if an interruption is considered a validated interruption.
At no time will multiple remedies be provided to the Customer for same, similar or related troubles for the same service. Customer payables must be current prior to the Customer receiving any credits
No credit shall be provided for interruptions due to the following:
the failure of any component, software, network or system provided by the Customer or a third party;
network maintenance by CloudatCost;
Force Majeure events;
acts or omissions, negligent or otherwise, of the Customer or a third party, or any breach of any term or condition by the Customer of the Master Services Agreement or any schedules or addenda attached thereto."http://www.cloudatcost.com/terms.php
edit: emphasis added
-
@scottalanmiller said:
Here is my IOWait from one of the boxes that just got brought back up...
01:53:57 PM all 12.47 0.00 9.12 26.66 0.00 51.75 01:55:07 PM all 23.65 0.00 14.91 37.25 0.00 24.2026% and 37%, those are crazy numbers!
02:20:01 PM CPU %user %nice %system %iowait %steal %idle
02:30:01 PM all 0.20 0.00 0.34 0.06 0.07 99.34
Average: all 0.20 0.00 0.34 0.06 0.07 99.34
^^SSD server with 2 CPU and 2GB of ram06:20:01 PM CPU %user %nice %system %iowait %steal %idle
06:30:02 PM all 0.28 0.00 0.15 0.69 0.03 98.84
06:40:01 PM all 0.36 0.00 0.20 0.82 0.05 98.57
Average: all 0.32 0.00 0.18 0.75 0.04 98.71^^SATA backup server with 512mb of ram and 1CPU
-
Just got the following via email:
Cloud At Cost System Message:
05/01/2015 05:36 PM
FREE CloudPro for all existing customers.
CloudatCost is offering for a limited time existing customers which have purchased servers using the "OneTime" option can migrate FREE to CloudPRO service with no monthly costs.CloudPro is the next generation of CloudatCost which will soon offer many extra features such as multi datacenter, redundant networks, additional OS Choices, advanced API and many more features scheduled for release in the next few weeks.
To migrate your existing server resources simply login to the panel at https://panel.cloudatcost.com, find your server you wish to migrate and select the "Migrate" option.
Please ensure you backup your data.
Once completed you can click CloudPRO and you will see all your resources you can start using.
-
Wasn't redundant networks one of the selling points of the current solution?
-
@scottalanmiller said:
Wasn't redundant networks one of the selling points of the current solution?
It's always been one of their selling points. That and using Fibernetics and their data center. They have redundant networks. So it was true. They just didn't fail over to them. So, It's honest marketing. We just assume redundant means they will fail over.
-
@thecreativeone91
Ya they have redundancy, but not automatic it seems.
-
@Sparkum said:
@thecreativeone91
Ya they have redundancy, but not automatic it seems.
Redudancy isn't automatic or not. It just is. It's the failover that's automatic. SAN sales people (and many others) pull this kind of stuff all the time.
