Geo- replication, and other new Windows Azure Storage.

Good day, friends! . Of course, not left unattended and Windows Azure. The conference devoted much time is Microsoft's cloud platform, and today we will talk about innovations in Windows Azure Storage.

Geo- Replication.

Customers continually emphasize the importance of disaster recovery capabilities in Windows Azure. Announced geo - replication mechanism helps in this area at no additional cost to the clients themselves. Geo- Replication can replicate data blobs and tables between the two data centers located in the same geographic region (eg between the data centers of northern and central America, or between the northern and western Europe ). So far, the mechanism does not replicate between two data centers located on different continents. At the moment, this mechanism is enabled by default for all existing accounts. I would also like to note that we do not feel any performance problems because the replication of a fully asynchronous.

New features of blobs, tables, and queues.

Updated version of the REST API and now he is the serial number, ... If the concrete is a list of new features as follows:.

• 
  Upsert - this is a new request to update the data in the table storage. Its peculiarity is that if the updated entity exists, then it is replaced by a new. If the entity does not exist, then create a new.
  


• 
  Projection tables - software allows customers to receive a subset of the properties of an entity. This improves performance by reducing the cost of serialization / deserialization, and reduces the load on the channel bandwidth.
  


• 
  New HTTP header for blobs - if specifically, have significantly improved downloading and streaming of blobs.
  


• 
  UpdateMessage - command allows you to put a lock on the message in the queue for the period of his treatment, as well as track the progress of processing the message.
  


• 
  InsertMessage - a new command that allows you to add new posts with a specific timeout. Message will remain visible in the queue until the specified timeout expires.
  

Let's take a little more in detail let's deal with the above new products. Thus, Table Upsert allows you to send a query or an update or addition to the effect, a concrete decision is made on the basis of its existence or absence. Naturally, this approach reduces the number of requests, since there is no need to pre- check the nature of its presence in the table.

Improved loading of blobs.

The header has been added, allowing to improve the download, streaming and download resumption of blobs from the repository. Previously, without the header, some browsers start downloading again, if for some reason the previous upload was interrupted.

UpdateMessage.

At the moment when the client reads the message from the queue it sets a timeout after which the client explicitly deletes a message, it will reappear in the queue. In many scenarios, the client needs to extend the timeout. Just for these scenarios and designed a new operation UpdateMessage. It allows you to control the visibility timeout thereby allowing the client to extend the timeout until the process is finished processing the message.
In addition, this operation allows you to update the post itself retained. This allows the client, processing the message, stored in a queue processing progress. And in case of unexpected failure, the other role can continue processing from the point at which the failure occurred.

InsertMessage.

In addition to the operation UpdateMessage has been added to the operation InsertMessage. It allows messages to be overlooked in the queue before the deadline, which allows us to plan future work by adding new posts, which become visible only after a certain period of time.
If you are interested in the above improvements, I recommend you read the following materials:.

• 
  Windows Azure Storage BUILD talk.
  


• 
  Geo-replication.
  


• 
  New Table Features.
  


• 
  New Queue Features.
  


• 
  New Blob Features.
  

Blogpost sources:. http://blogs. msdn. ru. aspx. ,. http://blogs. msdn. ru. aspx. ,. http://communities. quest. com/community/cloud_tools/blog/2011/10/22/geo-replication-for-windows-azure-storage.

Bayan, but nice

http://bablorub. ru / newyear / - Happy New Year, babloruby!.

Today the club for the prevention of mass change all passwords. Who has not come in the mail, first try to Generate a new. reminder. If you did not work, then please contact me at ICQ or e-mail, we shall understand, and will return access. Now login and password are case sensitive - be careful when entering.

news. Kazapy. : Taken by rail users in 6000, 2,000 projects and 42,000 outstanding options. Expected increase in turnover in cash in December - 20% of the November results.

Remember, you only have two days to overstocked. coupons for the purchase of software. at a discount. Hurry up time! .

Spomoni. asked to convey that he now writes a regular basis and invites everyone to his blog to try new materials and articles.

---------------------------------.
Sign up. to update the blog with one click!. Read on Twitter!. ::. Advertising blog.

Arguments ' for '

Not so long ago we held a round table discussion on SaaS, and invited him to the IT managers of insurance companies, banks and commercial structures. It was pretty comfortable, all this took place in Kyiv at the Opera in the evening, so it was an informal chat and a warm environment (thanks to Microsoft - all made ​​a fine).
so. According to the program were two of the report. One - our gene. Director about the use of cloud services and SaaS. It can look. here.

The second report, which wanted to talk in more detail - is the report of the IT director Valery Sokolov Generali Garant about how his company is shifting to services.
( Incidentally, this was an article in. Computer Active (N37). , So that you can read what and how. ).

In general, Valery - a man rather live in communion, and he spent his report in the format of a free narrative. Presentation of the result makes no sense, since the main value was exactly in the comments and responses to emerging issues. I'll try to briefly give the main points of interest.

a. Why did you decide to change the postal system?.
At the time of the decision the company has used e-mail program allows you to only work with mail. Western leaders came just did not understand such a scenario, because they are used to work with today's post, which would have included a calendar and contacts, and mobile access, etc. Dr.. It was therefore instructed to move to a normal system of corporate communications.

2. Why did you decide outsource?.
There are several reasons.
a) Because the crisis was on the one hand it is not clear how many people will be in the company of half a year - that is 2 times more, or less than 2-fold. T. e. It was impossible to predict the cost.
b) Guidelines adopted by the directive that IT should not turn the nut, and should solve the business problem.
c) have been counted the money, their value (t. e. how much the company would have earned if it had not spent the money to buy equipment and software, and would put them on deposit), maintenance costs, etc. Dr.. which resulted in a decent amount of.
d ) infrastructure in the company are not allowed to provide access to the system at an acceptable level - have been power outages, internet, etc. Dr..

3. Why is Exchange?.
Google and looked at the Exchange. They chose the latter for, quote, ' two and a half reasons '.
First, whereas in Ukraine was not Google's office ( now the truth is). What complicated the design of contracts, payments, etc. Dr..
Second, the company wanted to use the client software (outlook for example) and would like to achieve a zero- administration. T. e. so that users can install yourself and adjust it to the mailer. This was only possible using a bunch of outlook-Exchange ( which, incidentally, was done).
Second -half was used the following principle: to work with the service provider, which is smaller than you. In this case, we can expect that the supplier will try to satisfy a large customer to get money from it. Google in this case clearly does not fit.

4. Why SL Global Service?.
Considered Inkom and someone else. But at that moment (and at this point too), running a service with features such mastabirovaniya was only with us. We were given an opportunity to connect instantly and for free.

5. What about security?.
On this occasion, I wrote in. previous post.

6. What is the output?.
Well besides the fact that the company was expecting the worst, and got good service, there are several nice features:.
a) a decrease in dozens GPRS traffic and pay for GPRS roaming. Guide travels frequently and is now spending money in the fraction.
b) ustoychivat of service - Contrary to expectations, the IT department receives very few requests from employees and communicate with our support only if a real problem.
c) the fine-tuning spam filter - it was three days, when our guys a tie, then unscrew the filter level to achieve the ideal in terms of kilenta, your spam filter. The result is a well-.
d) removing the risks of security, maintenance, unplanned expenses, etc. Dr..

Well, in general, something like that. Again, listen to it was much more interesting. Sorry, did not record video - it was worth it.