Cut Over to New Data Center and Upgraded OS Done. :)


 
Thread Tools Search this Thread
The Lounge What is on Your Mind? Cut Over to New Data Center and Upgraded OS Done. :)
# 8  
Old 09-08-2018
I am not into VPS and clouds. We have tried them many times.

They are not reliable and perform poorly.

My experience is if you want reliability, spend the money and get a dedicated server in a well placed data center.

In my view "the cloud" is over-hyped, unreliable and very insecure.
This User Gave Thanks to Neo For This Post:
# 9  
Old 09-08-2018
Great job, Neo! Kudos for the smoothness of the transition.

Quote:
Originally Posted by Aia
Does it still make sense for you to be tied down to a private data center instead public cloud? I am curious.
Funny, i had just this discussion with a customer of mine: usually i work for banks and other large corporations with datacenters of several hundreds or thousands of (virtualised) servers. This time my customer is a relatively small company with about 50 servers (not virtualised by now) and the job is to transfer the whole datacenter to a new location (another town actually) and introduce virtualisation on the way.

Their first ideas when we had the kick-off work shop were the typical buzzword-bingo: cloud ... blabla ... software storage ... blabla ... Nutanics ... platform as a service ..., etc.

Then i asked a few questions: turns out, they have no idea what a "cloud" is and in fact what they really need is a reliable DevOps system. Their most prominent use case is: they need a test (development, ...) environment for one of their production servers and they want to more or less automatically deploy it. This is exactly what a DevOps system can help you with. My suggestion was to introduce Ansible but also reorganise their environment so that industrialisation and standardisation takes place. No more "hand-crafted one-of-a-kind" server systems but standardised machines with a standardised setup that is easily replicateable.

Next question: their most valuable asset is a large corporate database with a CRM system on top. Do you want to have that somewhere in the internet?
A: "Ah, but we can put a clause regarding this in the contract!"
Q: "Yes - and when this is breached (as has happened to some in the past) you get some refund but the trust of your customers to give you any data is NIL, never to return again and your business model is therefore dead. How much refund do you need to cover for that?"

Cloud services are nice when you need systems for testing for a limited amount of time and you don't want to hold the hardware resources for this. If you need a system to test your new database with meaningless test data then Amazon cloud, Microsoft Azure or whatever else you prefer is for you. If you need a reliable productive system you better have control over every clock cycle the underlying machine offers.

bakunin

Last edited by bakunin; 09-08-2018 at 05:49 PM..
This User Gave Thanks to bakunin For This Post:
# 10  
Old 09-08-2018
It is all about Continuous Integration and Continuous Delivery and Continuous Deployment. (CI/CD) If a customer does not see value on it, it might not be able to appreciate what the Cloud does bring, in general, whether a private one or a public one. Personally, I like engineering private cloud using as the base OpenStack and I have even done it using VMWare with VRA and VRO but ultimately the customer must understand the benefit of Infrastructure as Code and the safety that it produces by building confidence with testing and reproducible results.

Last edited by Aia; 09-08-2018 at 06:25 PM..
# 11  
Old 09-09-2018
Quote:
Originally Posted by Aia
It is all about Continuous Integration and Continuous Delivery and Continuous Deployment. (CI/CD) If a customer does not see value on it, it might not be able to appreciate what the Cloud does bring, in general, whether a private one or a public one. Personally, I like engineering private cloud using as the base OpenStack and I have even done it using VMWare with VRA and VRO but ultimately the customer must understand the benefit of Infrastructure as Code and the safety that it produces by building confidence with testing and reproducible results.

These are marking buzzwords, not engineering principles:


Code:
Continuous  Integration and Continuous Delivery and Continuous Deployment. (CI/CD)

Here is the problem.


Your post (on this topic, not generally) is one sided. This is normal and often seen when people have "drank the koolaid" and "believe in something" and develop a confirmation bias toward it. This is a fact of life - bias toward things we "believe in".


All good engineers (especially formally trained engineers) know from formal training and experience that every system has engineering trade-offs and it holds true, in general, that when you gain something in a system, you lose something else.


This is a universal truth.


So, all technologies, all systems, have good and bad, costs, benefits, negative and unforeseen consequences. This in an core universal constant. You cannot gain something without losing something.


Sorry to be so blunt, but when anyone, regardless of who they are or how smart or experienced they may be, make sweeping statements about now great something is without at the same time, making statements of the potential downsides - this demonstrates bias, not system engineering expertise or considered analysis.


Code:
Continuous  Integration and Continuous Delivery and Continuous Deployment. (CI/CD)


These are meaningless marking buzzwords designed by the marketing department of companies meant to sell and convince others to spend money.


I have a lot of experience in data centers (decades), and this means a lot of experience with both stand alone servers and cloud services.


While "cloud services" offer numerous benefits, they really do not offer the buzzwords above. To me, those buzzwords are more Orwellian than fact.


For example, if you have an application in a cloud with a lot of legacy code, and the cloud provider upgrades (which they do continuously as the buzzwords suggest) and your legacy code is not upwardly compatible with the upgrade of the service provider, you, the application owner, now have a problem which is driven by the cloud service provider, not by your own requirements, budget and timeline, etc.


This means that none of these buzzwords hold true in one simple test case, in practical terms when a cloud service makes changes to their infrastructure which is at odds with the customers legacy apps:


Code:
Continuous  Integration and Continuous Delivery and Continuous Deployment. (CI/CD)


In this case, these should be:


Code:
Continuous Upgrade Issues and Continuous Dependency and Continuous Trouble.


I can easily go on and on with experienced and well considered engineering trade-off examples and practical situational analysis to show that "cloud" technology has good and bad, value and devalue, and the cost-benefits are situational dependent.


If anyone does not understand this, and does not realize that all systems have positives and negative engineering trades, and tend to present one-sided arguments toward any system or technology, then that person is not yet a true engineer (in the core spirit of what an engineer is), but more of a "believer" or "follower". Engineers, especially system engineers, are not "believers" they are analysts who can see both sides of a system. System engineering know that every system is like the ying and yang system, with good and bad, and they think it looks foolish to espouse the positive without also pointing out the trade-offs.


On my desktop, for example, I have a 12 core Mac Pro. It's a great system (fast, strong, reliable, plenty of head room for all my apps and development IDEs to run); but this round black can puts out so much heat, my electricity bill has gone up. I need more cooling in the room because of it. So, I have to make a decision. Is the increased cost of electricity and heat generated at my desk worth the extra performance I get with 12 cores and 64G of RAM at my desktop? It's a trade-off.


So, now I automatically shut down this 12 core beast every night to reduce heat, which I never had to do with my little Mac Book Air!



Everything is a trade-off. Everything!


Marketeers, salespeople, and "followers" ("believers") tend to always only hype the positives, and hide the trade-offs. This is their job. Many people buy into the hype. Millions of people go out and make a living every day selling ideas. This is a core part of the economy, and without it, there would be less economy (another trade-off - hype does tend to make money and generate business).


This statement:


Code:
Continuous  Integration and Continuous Delivery and Continuous Deployment. (CI/CD)


Is pure "drink the koolaid" marketing buzz, and has nothing to do with sound system engineering analysis or trade-off analysis.


When folks reply to me, especially on broad concepts like this, please present both sides and not a one-sided "I am a believer" reply, because "believing" is really a kind illusion, at least in my viewpoint. Whenever we start to "believe" we start to lose objectivity.


Everything in this world is a trade-off, and everything has good and bad bundled together.


And if I can be so bold to say, it is these "strong beliefs" that people hold dear and near to their heart, which cause a great majority of the worlds problems.


I think I made my point and better stop now because I really go off the rails and start ranting against the current hype around AI !!
These 2 Users Gave Thanks to Neo For This Post:
# 12  
Old 09-09-2018
Quote:
Originally Posted by bakunin
Cloud services are nice when you need systems for testing for a limited amount of time and you don't want to hold the hardware resources for this. If you need a system to test your new database with meaningless test data then Amazon cloud, Microsoft Azure or whatever else you prefer is for you. If you need a reliable productive system you better have control over every clock cycle the underlying machine offers.

bakunin

Yes, that is certainly a good case for cloud services.


I also use cloud services for the same, but on a limited basis.


We used to run UNIX.COM in the cloud many years ago, but it was just a PITA compared to a top end dedicated box.


If we were not so "old and long in the tooth" here at UNIX.COM and did not have so much legacy PHP code, I might consider the cloud again, but we have far too much legacy code for cloud services.
# 13  
Old 09-09-2018
That feeling when everything is moved and working ...
Great job Smilie

As for the subtopic...
Quote:
RANT
'being in cloud' always meant not being realistic, imagining things in folks talk here Smilie

Promises of great things, but in practice.. unannounced downtimes, weird network issues, horror security and, of course, a human centipede / relay support.

This is for public clouds, for private implementation where you control all the nuts and bolts, you can make a nice and working environment.

I'm sick of that buzzword hype as well.

Went to some <insert vendor> conference.
They started talking about fog.
Bring your cloud closer, they said, with fog.

Took my a while to figure out those were proxies (squid for instance) with great caching capabilities in front of users, behind firewalls.
Well fog me stupid Smilie
/RANT
Regards
Peasant
These 2 Users Gave Thanks to Peasant For This Post:
# 14  
Old 09-09-2018
Quote:
Originally Posted by Neo
Your post (on this topic, not generally) is one sided. This is normal and often seen when people have "drank the koolaid" and "believe in something" and develop a confirmation bias toward it. This is a fact of life - bias toward things we "believe in".
Sorry, Neo, if I struck a nerve there, but I never said that CI/CD are engineering principles. I said that the Cloud is appreciated by those customers that implements CI/CD as part of their developing and delivering/deployment strategy. You had a lot to say and seems to have a strong opinion that anything opposite to your logic is "one sided". I do not have anything else to say.
Login or Register to Ask a Question

Previous Thread | Next Thread

5 More Discussions You Might Find Interesting

1. What is on Your Mind?

OUTAGE: Data Center Problem Resolved.

There was a problem with our data center today, creating a site outage (server unreachable). That problem has been resolved. Basically, it seems to have been a socially engineered denial-of-service attack against UNIX.com; which I stopped as soon as I found out what the problem was. Total... (2 Replies)
Discussion started by: Neo
2 Replies

2. What is on Your Mind?

Resolved: Issue in Server Data Center

Dear All, There was a problem in the data center data, which caused the server to be unreachable for about an hour. Server logs show the server did not crash or go down. Hence, I assume there was a networking issue at the data center. Still waiting for final word on what happened. ... (4 Replies)
Discussion started by: Neo
4 Replies

3. Shell Programming and Scripting

Failure rate of a node / Data center

Hi, Please, i have a history of the state of each node in my data center. an history about the failure of my cluster (UN: node up, DN: node down). Here is some lines of the history: 08:51:36 UN 127.0.0.1 08:51:36 UN 127.0.0.2 08:51:36 UN 127.0.0.3 08:53:50 DN 127.0.0.1 ... (6 Replies)
Discussion started by: chercheur111
6 Replies

4. HP-UX

Need to set up a HP cluster system in a data center

What are the server requirements, Software requirements, Network requirements etc, Please help me.. as 'm new 'm unable to get things done @ my end alone. Please refrain from typing subjects completely in upper case letters to get more attention, ty. (5 Replies)
Discussion started by: Sounddappan
5 Replies

5. Virtualization and Cloud Computing

Cloud Enabling Computing for the Next Generation Data Center

Hear how the changing needs of massive scale-out computing is driving a transfomation in technology and learn how HP is supporting this new evolution of the web. More... (1 Reply)
Discussion started by: Linux Bot
1 Replies
Login or Register to Ask a Question