Hi all,
I'm one of the software developers behind Cloud At Cost. I'll be available and representing myself during the Q&A in the upcoming week.
If you have any questions related to our panel as well as our new API (documentation is @ github.com/cloudatcost/api) feel free to shout out at that time or open up an issue within the Git repository.
@IRJ said:
@charlie Can you explain why there is a migration between cloud and cloudpro and exactly what that means? Is this a forced migration?
This is entirely a voluntary migration. We are temporarily allowing the migration of current life-time accounts to Cloud Pro as a means of thanking our customers for being patient with us during this outage.
If you are talking in regards to data. The reason for the need of the deletion / re-creation of the server is the infrastructure and underlining software that the cloud pro service runs on. In order to have a maintainable and sustainable service, we require a full deletion / re-creation of the service from our Cloud service to our Cloud Pro service.
@IRJ said:
@charlie Can you explain why there is a migration between cloud and cloudpro and exactly what that means? Is this a forced migration?
This is entirely a voluntary migration. We are temporarily allowing the migration of current life-time accounts to Cloud Pro as a means of thanking our customers for being patient with us during this outage.
If you are talking in regards to data. The reason for the need of the deletion / re-creation of the server is the infrastructure and underlining software that the cloud pro service runs on. In order to have a maintainable and sustainable service, we require a full deletion / re-creation of the service from our Cloud service to our Cloud Pro service.
Disclaimer: I work directly for CloudAtCost.
Due to the multiple Twitter messages that I have seen and received, I wanted to give some insight directly to Mango Lassi users in regards to the Cloud At Cost issues.
We did indeed experience a fiber cut and worked closely with our data center as well as Rogers in order to get this resolved. In response to @thecreativeone91 and @scottalanmiller I have included a picture of the accident that occurred and caused the outage in Waterloo, Ontario near Highway 7. One of our employees personally went there in order to get an insight as to why the disruption had been taking so long.
If you look closely at the picture, you can see on the right hand side the spool off Fiber cable that had been brought. At this point, Hydro One had yet to bring a pole to replace the one that had been damaged in the accident. I have additional pictures I can provide in regard to the outage if need be.
Coincidently, moments after the fiber line had been brought back up, there was a secondary outage in Oakville as confirmed by Rogers. At this point, we were programatically moving Cloud and Cloud Pro customers from the 10Gbit upstream link to our redundant link until the fiber link was 100% back up.
Going forward, we will be improving our redundant lines in order to sustain our current growth as well as network transfer rate.
As a thanks to our customers for being patient during this frustrating outage, we introduced the free migration from our Cloud service to our Cloud Pro service. Giving users the ability to convert their current life-time service to X,Y,Z resources that they may freely play with. For example, being able to convert a 4GB life-time VPS to resources and being able to deploy 8 512MB instances, or 2 2GB, 4 1GB, etc. This gives you the benefit of splitting one of your largers VPS' into multiple, scalable small VPS' on multiple different public IP's.
Due to the alarmingly positive response we've had from our Cloud Pro service, a considerable amount of customers have converted their current resources to Cloud Pro, thus our build queue has been considerably larger than usual. We have not only installed another SSD-storage to improve the build-speed, but also opened our gates to a few more build-servers in order to increase operations. The added benefit being that our overall I/O speed has improved across all servers. Check your VPS' I/O speed again and you'll see it has improved.
Currently we sit at 506 queued builds (building the config, expanding the /dev/sda, etc.) and 30 active installs (transfering data). We vary between processing 70-100 builds per hour.
Let me know if this is an adequate answer to your questions, we're all up for transparency and I'll be more than willing to give you more insight to our internal operations.
Hi all,
I'm one of the software developers behind Cloud At Cost. I'll be available and representing myself during the Q&A in the upcoming week.
If you have any questions related to our panel as well as our new API (documentation is @ github.com/cloudatcost/api) feel free to shout out at that time or open up an issue within the Git repository.