Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 5 years ago.
Improve this question
I'm comparing the cost of services like google cloud SQL to launching your own VM in the cloud with whatever sql version you'd like.
VM instance only vs Cloud SQL only
I'm quite surprised by the results I got. The cloud SQL, is more than twice the price for the same underlying system (8 vCPU, 52 Go RAM, 2 To storage)
So basically, you pay more to have less. And I was expecting the contrary...
Granted, you don't have to deal with maintenance and automating backup yourself, but I found the price difference ridicule.
So my question is : when should I consider using Cloud SQL instead of running my own specialized VM ?
Right now, I feel like this service is just a fancy way to milk money from the client.
Note : I took the google Cloud example, but this is the same result with other cloud providers.
The tl;dr answer here is that a VM is very different than a fully-managed service. It's like comparing apples and oranges, honestly.
When you create a VM, you have a VM. You can do whatever you want with it, but it's just a VM. That VM may be subject to restarts, must be totally configured by you in many cases, is not redundant, has no (added) security layer, etc.
As a managed service, Cloud SQL (and other managed services) offer many things way beyond what you can do on just a VM. You mention a fraction of them, such as backups. With a managed service you're getting a ton of other things which really matter to most people, such as:
Updates, upgrades
Better performance (in your example the IOPS of PD and Cloud SQL do not match)
Support for the service
Added security
An IAM layer
Integrations with other services
No need to "build it yourself"
etc...
While a (very) small minority of people may want to roll their own, it's generally a waste of time and a heck of a lot riskier than using a managed service. I think if you asked most any business customer, the cost of a managed service pales in comparison to paying a fleet of people to replicate the benefits you get from one.
This is true for GCP, AWS, and Azure.
Related
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 1 year ago.
Improve this question
Currently I'm using Azure Logic apps to sync the changes to different 3rd parties.
But it's too expensive when there are massive requests/messages.
The key features:
MQ connector, which can be used as trigger.
HTTP processor, used to issue HTTP requests.
Parse json response.
Possibility to check the history.
I've done some research of Apache Nifi.
My feeling is it's not very user friendly and quite old school.
One close open source option that I know of is n8n.
But you could also explore the fixed pricing model (Integration Service Environment) that logic apps offer, which is charged by the hour instead of based on the volume coming. Depending on the volume fluctuations, you can scale up more units as required.
Also, a completely new (in preview currently) way to develop and run logic app workflows was announced, which introduces a new pricing model (same as app service or premium plan of functions).
This is introduces a docker-based deployment which allows running your logic apps anywhere too.
Apache NiFi can be used for your requirements.
Apache NiFi is an easy to use, powerful, and reliable system to process and distribute data.
It has:
ConsumeMQTT processor which subscribes to a topic and receives messages from an MQTT broker.
InvokeHTTP processor which can interact with a configurable HTTP Endpoint.
Numerous json processors.
Data Provenance feature which tracks dataflow from beginning to end.
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 4 years ago.
Improve this question
As far as I am aware there is no DynamoDB equivalent database in transnational SQL land which is fully managed, as in there doesn't exist an SQL database which is implemented with a server-less architecture, every SQL database on AWS right now needs to be deployed to an EC2 instance.
Please correct me if I am wrong otherwise is there any theoretical reason why AWS or a third party does not create a server-less SQL database implementation? Or at least something close to a server-less implementation whereby there is no EC2 instance to manage?
Edit
Not sure what #aquinas or others are on about, none of the current RDS services are fully managed. IMPORTANT NOTE: It's actually also a question which sometimes comes up in your certification tests:
Which databases are fully managed?
Then you get a bunch of selections and only DynamoDB (atm see accepted answer) is fully managed, which is not a transactional SQL database. #aquinas is entirely incorrect, there is the notion of partially managed (as you would see in the explanation to that question if you were doing a practice AWS test when you get it wrong), in that you must specify when AWS should patch your database, how it should behave when it fails etc. Current RDS services don't have the elasticity as DyanmoDB and...
YOU HAVE AN EC2 INSTANCE???!!!?? screen shot attached!
However Amazon Aurora Serverless, brought to my attention from the accepted answer mentions in it's pricing section:
With Aurora Serverless, there are no instances to manage.
On a final note, these terms are not just semantics, they are important concepts in cloud computing. Fully Managed has a meaning, it's not open for interpretation as to what that means in AWS. Just because you use RDS to setup your EC2 instance running your database doesn't make it so. In this setup in the shared responsibility model means you are still controlling the EC2 instance, therefore meaning it is not fully managed!
You are right. Right now there is no service available on AWS.
But! Few weeks ago Amazon Aurora Serverless was announced and you are able to signup for preview access. So hopefully in the near future AWS supports a managed SQL solution.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 6 years ago.
Improve this question
Would you trust Jelastic if you were to use it for a SaaS application targeted at Enterprise customers?
I'm using it right now for its ease of use in creating a stack for a Java applcation. But it almost feels too easy.
What are you thoughts?
Jelastic is a universal product, designed and developed to satisfy the requirements of both expert and average users.
Ease and versatility are significant advantages of Jelastic. End user doesn't have to be a professional to use our product. But at the same time, Jelastic provides great opportunities for those people who use it.
If I needed a platform for SaaS application targeted at Enterprise, I would definitely choose Jelastic.
I cannot directly answer your question (since Layershift are a Jelastic hosting provider), but I can give you more background information that may help with your decision:
Jelastic is provided via a number of different hosting providers right around the world. That means you can combine the ease of use functionality that you mentioned, and use it together with the hosting provider that suits your needs best in terms of physical location, infrastructure details, support, SLAs etc.
Jelastic is also available in a "private cloud" model which allows you to use it either as a self-hosted on-premise solution, or remotely hosted at any datacentre (for example, any existing Jelastic hosting provider can offer this to you easily).
So there's a wide choice of possible solutions available to help you find the best option for your particular needs.
In contrast, IaaS options offered by AWS, Azure, Google etc. give you a single provider. You might like their platform but not their SLA options/prices etc., or you might like their SLA but not their platform - there is no diversity in their offerings.
My answer is NO.
For enterprise product not only we need to consider product itself we also need to take support quality into consideration.
I have been using Jelastic for almost two years and the biggest issues is support quality, the thing is Jelastic Hosting Partners are not trained with Jelastic knowledge, and the support team from Jelastic Hosting Partners can only act as middle man for Jelastic so the real support comes from Jelastic itself, the turnover time is too long for enterprise.
As time of writing, I already waited for my issue for 6 days, and still no response, come on, this is affecting my production!
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 2 years ago.
Improve this question
I need to store logs in a distributed file system.
Let's say that I have many types of logs. Each log type is recorded in file. But this file can be huge, so it must be distributed across many nodes (with replication for data durability).
These files must support append/get operations.
Is there a distributed system that achieves my needs?
Thanks!
I would recommend Flume, a log pulling infrastructure from the folks at Cloudera:
http://github.com/cloudera/flume
You can also try out Scribe from Facebook:
http://github.com/facebook/scribe
Combine a NAS with a no-sql database like MongoDB and you'll have distributed, large, and fault tolerant.
Of course, without more specific details like how much data, structure of the logs (or lack thereof), etc, it's really hard to recommend a real product.
For example, if by "huge" you really mean 2TB or less, and the data is highly structured, then a regular SQL server in a 2 machine clustered environment for fail over will do just fine.
However, if by "huge" you mean exabyte level or more and/or unstructured data then several large (and very expensive) NAS devices are needed. On which you run a set of no-sql databases that are clustered for fail/over and/or multi-master relationships...
You can use Logstash to collect the logs and centralize them with an Elasticsearch cluster. The local logs could be rolling log files, so that they remain small.
Further you can use Graylog2 to analyze and view your logs.
Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.
We don’t allow questions seeking recommendations for books, tools, software libraries, and more. You can edit the question so it can be answered with facts and citations.
Closed 2 years ago.
Improve this question
I need to Track Unique Visitor count in my web application. I would really like to use Google Analytics but due to the Load limitations that google imposes I will not be able to use them. I am expecting WAY over 10,000 requests a day. This is the limitation that Google web analytics API imposes. Is there another company that has the same features as google analytics that is paid or free?
There definitely are.
Here are two open source and free solutions that are very polished:
Piwik - Designed as a direct competitor to Google Analytics (it looks just as nice) that you host on your own servers
Open Web Analytics
the 10,000 request apply to the Data API, not to the actual data collection.
Like you can have an unlimited number of users seeing your website. On the other hand if you use the API to extract data from their database, you can do 10k request a day only.
check this link for more details
The biggest, most obvious, most usual alternative is to simply do it yourself. Your webserver needs to log requests for security etc. anyway, so it's not a big deal to run something like webalizer on those logs. You won't get the quick, easy access to advanced information like paths users take through the site, btu that can be determined if you care enough. You do gain one huge benefit though: privacy of your own data.
We use Omniture here but it'll cost you.
There is SpeedTrap, a java-based analytics package. Our company used it for years before they turned into cheap **ards and decided Google Analytics was more cost effective (because it was free). But that's a story for another night.