Finally the help of IT is here

Blog of computer solutions.

As the CPU operating in Amazon AWS Credits

Posted by Xaus Xavier Nadal 19th on August 2015

The Credits CPU Amazon AWS are responsible for increasing CPU performance of t2 instance in case of specific need, t2 instances also calls burstable performance or burst have a fixed defined performance as we can see in the following table and extra temporary capacity limited by the CPU Credits defined.

This means that when hiring a type EC2 accept t2 instance that does not have a fixed return over time and can vary depending on whether the CPU credits have been achieved or not to get 100% CPU performance.

AWS CPUCredit_0

The instance starts with a credit balance sufficient CPU for the performance is good, while this instance in idle mode (or without using your baseline performance) the CPU Credits accumulate from the table above CPU Credits / hour, Up to 24 hours, where if you have not spent will accrue not.

For an instance it requires a level higher than its CPU Baseline Performance at peak times (burst) will use accumulated credits available to complete them and continue with the performance marked their baseline performance

There is no differentiation between instances use Windows or Linux.

AWS CPU Credit

CPU credits can be monitored thanks to the service CloudWatch to detect whether we need to increase the instance type t2 and have more credits CPU or start thinking about moving to an instance fixed performance as m3, c3, r3 instances ...

Let's take an example to make it clearer.

Suppose we have an instance t2.micro 10% lit using a continuous CPU (Its maximum allowed) in this case are accumulated at a rate of CPU 6 credits every hour. Passed 2 hours we will have 12 minutes full (2 hours x 6 credits) in which we can use the 100% of performance of a core CPU at maximum performance, after these 12 minutes the system will lower your performance up to 10% power the CPU core that is assigned to your baseline.

NOTE: The operating system will indicate that it is the 100% CPU if the process is not complete, but go a 90% slower than in 12 minutes before where we could enjoy the full power of the physical core assigned to the virtual machine.

For an instance t2.large always will have a 60% of core CPU for our use and enjoyment, in the specific case we need to overcome this performance, the system will catch your CPU reservation Crédits the accumulated rate of 36 credits per hour set in the first image from the article.

If the CPU usage of the t2.large instance is less than one 60% of CPU during 24 hours, we would have 14 864 hours or minutes of use of the system to 100% of its processing capacity (24 36 hours * credits), after this time would have a 60% yield of the core.

Remember that the internal performance of the operating system never displays the actual CPU usage of the instance because the percentage of CPU usage is on the use of credits based CPU and performance of each instance.

Since I know what you're going to ask and I answer directly.

In the case of medium and large t2 instances that have 2 vCPUs means they have twice the capacity of process? Well, no. In the case of t2 medium is based on a 40% of a core available for a single process (1 vCPUs) but using multiprocessing (2 vCPUs) have a 20% for each process. in the case of t2.large the same but with a vCPU 60% and a 30% if multithreaded.

As it is a very complicated issue to understand, I leave the comments box so that you can review and submit your contributions are always welcome.

MegaCracks greetings.

tags: , , ,
Posted by Amazon AWS | 1 Comment »

Enable Memcached on CentOS with Putty

Posted by Xaus Xavier Nadal 22nd on July 2013

Goodnight.

Today we will see how step install step Memcached on CentOS, also will serve for newbies in the field for your hosting, we will in a very generic way and can be used on any virtual server (VPS) or dedicated server or even in your shared hosting.

We're going to do with putty for which we have previously set up an SSH connection from our web server. What will we do with root credentials (perhaps missing unto you these credentials) but if you have got them this is your step article.

After connecting with Putty we will ask the server a login and password, put the of your root.

The first thing we do is create a temporary directory with the mkdir command which will download memcache in its stable version in the example of 22 / 9 / 2012 2.2.7 version.

mkdir repo

repo cd

Now we will download from the Internet repository the latest stable with the following command version.

wget http://pecl.php.net/get/memcache-2.2.7.tgz

Memcached 1

As you see in the picture we have made a ls to see that the file has been downloaded on this route.

 

 

Megacrack read »

tags: , , , ,
Posted by web performance | No Comments »

15 plugins for WordPress (SEO, performance, outgoing)

Posted by Xaus Xavier Nadal 12th on October 2012

The following article shows the 15 plugins for WordPress that we are using in this blog.

Many blogs show interesting plugins in articles but no one had ever shown so far plugins that are used in the blog.

I think it's important the latter as it gives confidence to the reader that these plugins for WordPress are really tested and recommended.

I know it's a security breach exposing WordPress plugins we are using in the blog since any failure in the code of these can cause us to be vulnerable to attacks, I hope it's not so for the sake of the followers of the blog, but for this we are also careful to keep the blog with the latest security updates.

I hope you like and we comenteis WordPress plugins that you use to you in your blogs.

First I'll show them all at once with its links to the link to download and then discuss each of them separately.

Akismet

Authors Widget

Dagon Design Sitemap Generator

FeedSmith FeedBurner

Google XML Sitemaps

GTranslate

Breadcrumb RDFa

Real IP

ShareThis

W3 Total Cache.

WordPress SEO

WP-Optimize

WP Smush.it

WPtouch

Yet Another Related Posts Plugin

Megacrack read »

tags: , , , , , ,
Posted by general | No Comments »

Solve problems Outlook and Exchange Performance

Posted by Xaus Xavier Nadal 10th on May 2012

Good Morning.

For some years that Outlook users with Exchange 2003 are experiencing slow systems, time out, errors, etc .. Each day the amount of mail we receive is higher, the size of e is higher and capabilities of servers they remain the same (subject crisis).

But the mistake that the mail server is responsible, is a wrong way to take on the problem.

Outlook 2003 / 2007 / 2010 like many other mail systems have their limitations and best practices that should follow strictly and only without modifying the Outlook system users and especially without investing in new hardware or even without thinking about upgrading your system mail to a recent version we could achieve very good results in performance mail system is concerned. (The user will thank you.)

One of the major problems and very easy to solve is the number of emails in each folder in Outlook.

This number post should never exceed 3000 post (Sounds like a lot at first, but look at your inbox or sent items folder (How many have ???).

With this simple change we can see a substantial improvement in the performance of our Outlook and therefore in the user feedback regarding the mail system. And if you look also amazing on the Exchange server load.

The contents of each folder is stored in a table in the Information Store Database, as it increases the number of items in each folder mechanism Extensible Storage also called ESE using data structures B + trees to store records also increases and as the number of records increases, the number of I / O disk to store new information also increases and decreases the performance of the entire system.

Megacrack read »

tags: , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,
Posted by Exchange | No Comments »

Migration to faster server (got 100 / 100 web performance)

Posted by Xaus Xavier Nadal 3rd on May 2012

Good,

Today maybe for a while you've not been able to enter the blog, but the interruption was for the better.

Sorry I could not tell before but we had a core operating system upgrade of our hosting provider Bluehost.

This morning we received the following letter.

Dear Xavier,

 

We're pleased to inform you That the server hosting your website will be megacrack.es Undergoing a major software upgrade, from CentOS to CentOS 5 6, 48 Within the next hours.

 

This upgrade includes newer software packages (including Python, Perl and gcc), as well as all the security and performance benefits That eats 6 CentOS Along With. In Addition to this, the server will be redeployed with a different file system type Simultaneously, further increase increasing performance.

 

, Although a bulk of the upgrades to your server are being done online and functional With It, in order to safely finalize These Changes will need our Administrators to temporarily take your server offline in the early morning hours. Barring any extenuating circumstances esta outage last only About 2 Should hours.

 

Please note That while we do not anticipate your software-update post Having problems, it May be required to re-compile any module (s) you are using to take advantage of the newer included libraries. We Suggest reviewing your site afterward to verify it is That Functioning as it should.

 

If You have any questions or please do not hesitate Concerns to contact us.

The truth is I've done a performance test and finally have a First Byte Time de 506 ms. Bieeeeeeeeennn !!!!! Megacrack read »

tags: , , , , , , , , , , , ,
Posted by Improving Web Speed, Web | 1 Comment »