AWS AMAZON - how you optimize resources

Good day!

I have been using AWS AMAZON for a long time when it is necessary to increase the rated power. Satisfied, well, in general, everything. In short, both permanent and spot servers were launched for tasks. Various types of instances and auto start / stop schedules were used.
In general, all this is very comfortably regulated, to a first approximation, with up to a dozen servers and with my direct understanding of the whole kitchen. And of course, accounting, export / archiving / disposal of stored data.

I wonder what is your “gentleman's set” for optimizing resources and, ultimately, costs?

Thank you for your comments, recommendations!

...

The question is, for managers who manage more than a hundred servers *, of which at the same time, for example, the same dozen can be launched, but each server has its own “ballast” from snapshot images, etc. paraphernalia. Elastic IPs that migrate / can be “bound” as permanent. And all this needs to be known (to know where to look).
All the same, size matters, and if at “10” the launch of the servers during the user's working hours (up to ~ 60%) brings significant savings / data replication to a more powerful spot for fast data processing and subsequent termination into nonexistence. That is for “100+” this is a more capacious question.

Perhaps you implemented AMI / Snapshot storage in Glacier, or are there tricky schemes? The question, by the way, is very interesting - if I am not mistaken in Glacier, you can store archives / data located outside the "AWS console", and the idea to feed for 0.01 in Glacier AMI is very good.

I would like to think about the prospect of expansion, but division of labor is also possible, it was not without reason that IAM was implemented. And if you were given (planned) a certain “100+” server pool with its own ballast?

... Or "Elasticfox", scripts and tables + drill, in terms of receiving information from customers, what would then dock the tails?

Or maybe, in general, it makes cardinal sense to move to competitors for permanent residence, to minimize costs, and this is already a transfer task and should be justified significantly, can anyone have experience?

A pair of tables, only RAM and CPU were compared, attracting AMAZONA prices by the ears.

Clouds overseas
TypeCloudname (instance-types)RAM GiBCPUWindows Usage (per Hour)
Swin azureSmall (A1)1.701$ 0,090
SAwsSmall instance1.701$ 0,091
ShpcloudSmall2.002$ 0.120
Srackspace2 GB2.002$ 0.120
Ssoftlayer2 Core + 2GB RAM2.002$ 0.250
Mwin azureMedium (A2)3,502$ 0.180
MAwsMedium instance3.752$ 0.182
MhpcloudMedium4.002$ 0.240
Mrackspace4 GB4.002$ 0.240
Msoftlayer4 Core + 4GB RAM4.004$ 0.390
Lwin azureLarge (A3)74$ 0.360
LAwsLarge instance7.504$ 0.364
Lsoftlayer4 Core + 8GB RAM8.004$ 0.440
LhpcloudLarge8.004$ 0.480
Lrackspace8 GB8.004$ 0.480


Clouds domestic:
TypeCloudname (instance-types)RAM GiBCPUWindows Usage (per Hour)
Sselectel.ruSmall1.701$ 0,063
Soversun.ruSmall2.002.6$ 0,070
SAwsSmall instance1.701$ 0,091
Sscalaxy.ruSmall1,5041$ 0.155
Mselectel.ruMedium3.752$ 0.132
Moversun.ruMedium42.6$ 0.180
MAwsMedium instance3.752$ 0.182
Mscalaxy.ruMedium4.0041$ 0.321
Lselectel.ruLarge7.504$ 0.264
Loversun.ruLarge85.2$ 0.323
LAwsLarge instance7.504$ 0.364
Lscalaxy.ruLarge8.0041$ 0.588

Data is sorted by type and price.
As you can see from the tables, only the “Windows” servers were compared, since they make up 99% in my park.

PS:
I apologize in advance for the multi-letter, but the question, “What is your“ gentleman's set ”” in AWS, somehow looks like a heap.
* Of course, all useful opinions are interesting, “100+” is a kind of figurative value.

Also popular now: