I'm a CS student, just exploring the SCM space. While doing my own research I came across many different hosted solutions (GitHub obviously, Lighthouse, YouTrack, TeamCity, etc.) - do you think it is actually reasonable to try to host a (commercial, closed source) project entirely in the cloud?
Let's say I'd host code on GitHub, use Jira or Lighthouse for issue tracking, God knows what other hosted PM solution (Basecamp?) and build using EC2 (I can put Hudson or TeamCity on it and use appropriate EC2 plugins for these products to get more computing power when needed).
Is the EC2 bill going to kill me (compared to self-hosted solutions)? Do you think "the cloud" it's still not reliable enough?
This is the way we work at our company. Version control system (git) + agile planning + ticket system/bugtracker + wiki are hosted at http://www.assembla.com for 49$/month for 40 users, private repositories ( https://www.assembla.com/plans ) and we have a micro instance on amazon aws ec2 where jenkins, nexus, sonar and some internals tools are running for free the first year and then you should consider spending like 80$/month for the same service.
So it costs 129$/month for a full cloud solution for a small company (40 users max): reliable, with a good release train of new features by our service providers and with a low maintenance footprint for us.
Compared to self hosted it's not really expensive considering the following costs :
- price of your server (lets say 1000$)
- electricity bills (lets say 30$/month for 100% uptime)
- cost of configuration (to get the same quality as assembla for exemple) and maintenance (lets say 0.5day man per month at 500$/day in france)
The cost is : 363$/month
This should look a bit biased, but finally it's what we experienced.
Regards,
Xavier
There is no problem to use the cloud for hosting, and many large companies do so already. I think NetFlix recently moved solely to EC2. Our whole business runs on EC2, and it's been relatively good so far.
The EC2 bill is up to you to manage -- cloud is all about granular billing for services, and the more you consume the more you pay (we sell a tool that helps with cost controls: http://LabSlice.com). Your biggest cost will usually be CPU power, so stick to the Micro/Small instances until you've got a handle on costs.
It's interesting that people question the reliability of cloud, as the underlying premise is actually to provide more reliability to businesses then they could afford themselves (high scalability, immediate availability of hardware, monitoring, load balancing etc.)
You can make use of AWS Free Account and host your application. If you exceed Free Account Usage limit,you will be charged for whatever extra you have utilized.
Regarding reliability about Cloud, every big firm is moving towards Cloud like Amazon,Microsoft,IBM,HP etc because they found cloud reliable,cost effective and green.
Given your a student and assuming your looking to spend little money, a lot of Git and SVN hosting providers offer free hosting for students or free accounts if your a small team with minimal storage requirements. Check out Codesion's student offering for example (disclaimer, I work for Codesion). This plan also comes with Trac / Bugzilla for your PM requirements. I wouldn't be concerned with security and reliability for the same reason that Simon points out above.
As for CI on EC2 - this is probably your best bet since you pay by the hour each instance is running. I'd recommend using Amazons API to fire up an instance each time Hudson needs to perform a build, store the results of the build on more permanent storage, and shut the instance down when finished. If your doing a lot of CI builds, it might be better to just keep the instance running, but this will cost you more of course.
Related
On Db2 v11.1, how do we get or setup the notification for DBA team if there is any hang or slowness situation in off shift working hours?
The answer depends on the external monitoring and alerting solution you deployed, and how you configure that tooling in your environment.
This application layer tooling is not built into Db2-LUW, although APIs exist in Db2-LUW for such tooling to get the data it needs in order to operate.
IBM and several third parties offer solutions for real time monitoring and alerting in this space. Many cover app-servers, web-servers, database layers, networks and operating-system layers and have different alerting configurability. Many have plugin type architecture with plugins for Db2-LUW monitoring. Do not use stackoverflow for product recommendations however.
For "slowness", this is only meaningful to measure usually at the application layer, in terms of response times and other metrics etc.
For database-hangs, IBM offers a db2-hang_detect script that tooling can orchestrate , requires careful interpretation and even more careful testing.
I am new to Kubernetes.
Last few days, I did performance test for API Gateway to compare both environments, Kubernetes and Native.
I expected to get similar results in both types of tests, but the second test showed a 20% tps loss.
How to stress : RestfulAPI requests through API Gateway.
Env 1 : Native (CentOS) - result : approximately 1700 tps
Evn 2 : Kubernetes - result : approximately 1350 tps
* Hardware specifications were the same.
I wonder if this is a common result.
In my opinion, 20% is too much loss.
Sorry for my stupid question, but I am confused.
According to k8s documentation
Kubernetes is a portable, extensible, open-source platform for managing containerized workloads and services, that facilitates both declarative configuration and automation. It has a large, rapidly growing ecosystem. Kubernetes services, support, and tools are widely available.
The name Kubernetes originates from Greek, meaning helmsman or pilot.
Wikipedia also says:
Kubernetes (κυβερνήτης, Greek for "governor", "helmsman" or "captain")
Obviously, the captain’s advantages are not in speed but in the predictability of the work of the entire complex system as a whole
k8s strengths are:
Service discovery and load balancing
Storage orchestration
Automated rollouts and rollbacks
Automatic bin packing
Self-healing
Secret and configuration management
As you can see, the performance gain is not mentioned here.
P.S. There is great comment on HN discussion about Serverless hype (TL;DR: slowness and costiness):
It's concerning how typical the hype machine is in IT. I believe Serverless has its place and value. So does Kubernetes or many other products that are often discussed on HN.
But let's be clear, we are talking about commercial products and there is a capital interest in selling these services to all of us devs and engineers.
So while use cases exists and benefits wait to be reaped, as a consultant I strongly feel that we should be MUCH more insistent in pointing out when a product does not make sense instead of jumping onto the hype train.
I am basically surrounded by "LETS TRANSFORM EVERYTHING TO KUBERNETES THIS WEEK!" exclamations, conferences are basically "DID YOU ALREADY HEAR ABOUT KUBERNETES?" and so on ...
It reminds me of Ruby on Rails, a mature and well-developed framework used by global tech firms (correct me if wrong: Airbnb, ~Stack Overflow~, Github) to handle parts of their backend in 2019. But for half a decade now even tiny companies have been screaming around about FancyHTTPTechThisYear (tm) because scale while reporting 1/500th of the traffic of some famous Rails users.
This is not engineering with objectives in mind, it's more akin to the gaming community yelling for a new console.
our small team of 2 people recently got upgraded to 5, which means we should introduce a bit more infrastructure around our project in order to work efficiently together. Its a university research project.
How much administrative effort is it to run
Gitlab
Jenkins
A Release Server
on a rented machine, in comparison to a SaaS solution e.g github with Travis ?
Unfortunately nobody in our (quite academic) team has practical experience with that. I know the setup can be done fast, but how time consuming is it to keep the stuff running. Are there other concerns we might be missing?
Of course we would like to mainly work on the project itself, but since the toolstack keeps growing by time we are not sure if a SaaS solution is what we need.
With these inputs be ready to spend from one to several weeks of one from the five.
Instead you might want to check which leading hosting companies provide free services for open source or educational projects.
I was involved in couple of cloud computing platform recently.
First of all please note that I am not trying to criticize any platform.
Cloud computing is large area but to make my point simple and understandable. Let me come up with very simple scenario and that is data storage services hosted on the cloud.
If you take any storage service like Amazon EC2, SQL Data Service(SDS), Salesforce.com services.
If you want to consume any of such data storage service platform goal of all such service are same and that is to serve requested data on demand. Without warring about how it store and where it stored and who is maintaining it etc... (all cloud goodies)
Now my area of concern is the way ANSI-SQL regulated platform venders to make sure they follow similar language across all the product can’t they regulate similar concept across
service providers?
Why no such initiatives??
Any thoughts appreciated
It seems to me like you're worried about vendor lock-in with cloud computing. I may be naive but I would normally choose technologies and then go look for cloud vendors that'd be able to deliver these technologies. And if I was aiming for a "write once run anywhere approach" I'd have to select technology that'd make this as realistic as possible.
With the fairly rapid speed of development I really think standardization committees would struggle to keep up. ANSI-SQL has had 20 + years of history. It seems to me like you're requesting for standardization long before we even know what the cloud is up to....
I think that this emerging cloud computing initiative is just too young in order to have standards.
Service providers right now just worry about rushing into the market, rather than interoperability and standards.
Later on, when the situation is more established, some common guidelines may emerge. But there is still a long way to go.
You seem to be asking specifically about cloud storage services, rather than cloud computing in general. So your Amazon example would be S3, not EC2.
I think the field is a little young to be standardising on an API just yet. The services differentiate themselves in ways which rule this out. For example, S3 trades sophistication for scalability/reliability/performance: you can't do a complex SQL LIKE query. You can store and retrieve blobs of data based on a key, and that's about it.
I think as such services become more and more the mainstream way to do things, standards will emerge. Users will want the freedom to switch providers on a whim, move their data around, test against free local storage, etc.
The APIs used are all based on Web Standards already. Making an abstraction layer to make them look the same is fairly trivial.
I'm looking for a way to give out preview or demo versions of our software to our customers as easy as possible.
The software we are currently developing is a pretty big project. It consists of a client environment, an application server, various databases, web services host etc.
The project is developed incrementally and we want to ship the bits in intervals of one to two months. The first deliveries will not be used in production. They have the puropse of a demo to encourage the customers to give feedback.
We don't want to put burden on the customers to install and configure the system. All in all we are looking for a way to ease the deployment, installation and configuration pain.
What I thought of was to use a virtualizing technique to preinstall and preconfigure a virtual machine with all components that are neccessary. Our customers just have to mount the virtual image and run the application.
I would like to hear from folks who use this technique. I suppose there are some difficulties as well. Especially, what about licensing issues with the installed OS?
Perhaps it is possible to have the virtual machine expire after a certain period of time.
Any experiences out there?
Since you're looking at an entire application stack, you'll need to virtualize the entire server to provide your customers with a realistic demo experience. Thinstall is great for single apps, but not an entire stack....
Microsoft have licensing schemes for this type of situation, since it's only been used for demonstration purposes and not production use a TechNet subscription might just cover you. Give your local Microsoft licensing centre a call to discuss, unlike the offshore support teams they're really helpful and friendly.
For running the 'stack' with the least overhead for your clients, I suggest using VMware. The customers can download the free VMware player, load up the machines (or multiple machines) and get a feel for the system... Microsoft Virtual PC or Virtual Server is going to be a bit more intrusive and not quite the "plug n play" solution that you're looking for.
If you're only looking to ship the application, consider either thinstall or providing Citrix / Terminal services access - customers can remotely login to your own (test) machines and run what they need.
Personally if it's doable, a standalone system would be best - tell your customers install vmware player, then run this app... which launches the various parts of your application stack (maybe off of a DVD) and you've got a fully self contained demo for the marketing guys to pimp out :)
You should take a look at thinstall(It has been bought by vmware and is called thinapp now), its an application virtualizer.
It seems that you're trying to accomplish several competing goals:
"Give" the customer something.
Simplify and ease the customer experience.
Ensure the various components coexist and interact happily.
Accommodate licensing restrictions, both yours and the OS vendor's.
Allow incremental and piecewise upgrades.
Can you achieve all of these by hosting the back end (database, web server, etc.) and providing your customers with a CD (or download) that contains the client? This will give them the "download/upgrade experience" that goes along with client software, without dealing with the complexity of administering the back end.
For a near plug-and-play experience, you might consider placing your demo on a live linux or Windows CD. Note: you need a licensed copy of Windows for the latter.
Perhaps your "serious" customers might be able to request their own demo copies of the back end as well; they'd be more amenable to the additional work on their part.
As far as OS licenses, if your vendor(s) of choice aren't helpful, you might consider free or open-source alternatives such as FreeDOS or linux.
Depending on if you can fit all the needed services into a single OS instance or not...
Vmware Ace or whatever they're calling it nowadays will let you deliver single virtual machines under strict control, with forced updates, expiration and whatnot. But it sounds easier to just set up a demo environment and allow remote access to it.
The issue here I guess is getting several virtual machines to communicate under unknown circumstances - if one is not enough?
An idea then is to ship a physical server preconfigured with virtualisation and whatever amount of virtual servers needed to demonstrate the system.
Using trial versions of the operating system might be good enough for the licensing dilemma - atleast Windows Server is testable for 60 days, extendable to 240 when registering.
Thinstall is great for single apps, but not an entire stack....
I didn't try it yet, but with the new version of thinstall you are able to let different thinstalled application communicate.
But I guess you're right a vm-ware image would be easier