We are finding Azure SQL Database very slow. Its probably about 10x slower than the same spend on a SQL on an Azure VM.
However a VM based solution requires maintenance and backups and im concerned I'll loose the VM and data if something horrible goes wrong.
Thus the SQL Azure solution seems safer to me. So I have two specific questions.
Are you seeing this speed difference and if so is there a solution
Is there any nice solution for ensure the SQL on VM is backed up automatically and offsite.
Azure SQL database is slightly slower than SQL Server on Azure Virtual Machine. However, we didn't find it as slow as 10X. May be you should try Premium tier, which delivers more powerful and predictable performance, in case your database is in some other tier.
Regarding SQL Server on Azure Virtual Machine, there is support available for automated backup and patching. Please visit the below link for more details.
http://azure.microsoft.com/blog/2015/01/29/automated-everything-with-sql-server-on-iaas-vms/
While performance analysis, it is worth checking for ASYNC_NETWORK_IO wait type. Make Sure your client application is deployed close to your database in Azure.
Related
I'm currently running an instance of MS SQL Server 2014 (12.1.4100.1) on a dedicated machine I rent for $270/month with the following specs:
Intel Xeon E5-1660 processor (six physical 3.3ghz cores +
hyperthreading + turbo->3.9ghz)
64 GB registered DDR3 ECC memory
240GB Intel SSD
45000 GB of bandwidth transfer
I've been toying around with Azure SQL Database for a bit now, and have been entertaining the idea of switching over to their platform. I fired up an Azure SQL Database using their P2 Premium pricing tier on a V12 server (just to test things out), and loaded a copy of my existing database (from the dedicated machine).
I ran several sets of queries side-by-side, one against the database on the dedicated machine, and one against the P2 Azure SQL Database. The results were sort of shocking: my dedicated machine outperformed (in terms of execution time) the Azure db by a huge margin each time. Typically, the dedicated db instance would finish in under 1/2 to 1/3 of the time that it took the Azure db to execute.
Now, I understand the many benefits of the Azure platform. It's managed vs. my non-managed setup on the dedicated machine, they have point-in-time restore better than what I have, the firewall is easily configured, there's geo-replication, etc., etc. But I have a database with hundreds of tables with tens to hundreds of millions of records in each table, and sometimes need to query across multiple joins, etc., so performance in terms of execution time really matters. I just find it shocking that a ~$930/month service performs that poorly next to a $270/month dedicated machine rental. I'm still pretty new to SQL as a whole, and very new to servers/etc., but does this not add up to anyone else? Does anyone perhaps have some insight into something I'm missing here, or are those other, "managed" features of Azure SQL Database supposed to make up the difference in price?
Bottom line is I'm beginning to outgrow even my dedicated machine's capabilities, and I had really been hoping that Azure's SQL Database would be a nice, next stepping stone, but unless I'm missing something, it's not. I'm too small of a business still to go out and spend hundreds of thousands on some other platform.
Anyone have any advice on if I'm missing something, or is the performance I'm seeing in line with what you would expect? Do I have any other options that can produce better performance than the dedicated machine I'm running currently, but don't cost in the tens of thousand/month? Is there something I can do (configuration/setting) for my Azure SQL Database that would boost execution time? Again, any help is appreciated.
EDIT: Let me revise my question to maybe make it a little more clear: is what I'm seeing in terms of sheer execution time performance to be expected, where a dedicated server # $270/month is well outperforming Microsoft's Azure SQL DB P2 tier # $930/month? Ignore the other "perks" like managed vs. unmanaged, ignore intended use like Azure being meant for production, etc. I just need to know if I'm missing something with Azure SQL DB, or if I really am supposed to get MUCH better performance out of a single dedicated machine.
(Disclaimer: I work for Microsoft, though not on Azure or SQL Server).
"Azure SQL" isn't equivalent to "SQL Server" - and I personally wish that we did offer a kind of "hosted SQL Server" instead of Azure SQL.
On the surface the two are the same: they're both relational database systems with the power of T-SQL to query them (well, they both, under-the-hood use the same DBMS).
Azure SQL is different in that the idea is that you have two databases: a development database using a local SQL Server (ideally 2012 or later) and a production database on Azure SQL. You (should) never modify the Azure SQL database directly, and indeed you'll find that SSMS does not offer design tools (Table Designer, View Designer, etc) for Azure SQL. Instead, you design and work with your local SQL Server database and create "DACPAC" files (or special "change" XML files, which can be generated by SSDT) which then modify your Azure DB such that it copies your dev DB, a kind of "design replication" system.
Otherwise, as you noticed, Azure SQL offers built-in resiliency, backups, simplified administration, etc.
As for performance, is it possible you were missing indexes or other optimizations? You also might notice slightly higher latency with Azure SQL compared to a local SQL Server, I've seen ping times (from an Azure VM to an Azure SQL host) around 5-10ms, which means you should design your application to be less-chatty or to parallelise data retrieval operations in order to reduce page load times (assuming this is a web-application you're building).
Perf and availability aside, there are several other important factors to consider:
Total cost: your $270 rental cost is only one of many cost factors. Space, power and hvac are other physical costs. Then there's the cost of administration. Think work you have to do each patch Tuesday and when either Windows or SQL Server ships a service pack or cumulative update. Even if you don't test them before rolling out, it still takes time and effort. If you do test, then there's a second machine and duplicating the product instance and workload for test.
Security: there is a LOT written about how bad and dangerous and risky it is to store any data you care about in the cloud. Personally, I've seen way worse implementations and processes on security with local servers (even in banks and federal agencies) than I've seen with any of the major cloud providers (Microsoft, Amazon, Google). It's a lot of work getting things right then even more work keeping them right. Also, you can see and audit their security SLAs (See Azure's at http://azure.microsoft.com/en-us/support/trust-center/).
Scalability: not just raw scalability but the cost and effort to scale. Azure SQL DB recently released the huge P11 edition which has 7x the compute capacity of the P2 you tested with. Scaling up and down is not instantaneous but really easy and reasonably quick. Best part is (for me anyway), it can be bumped to some higher edition when I run large queries or reindex operations then back down again for "normal" loads. This is hard to do with a regular SQL Server on bare metal - either rent/buy a really big box that sits idle 90% of the time or take downtime to move. Slightly easier if in a VM; you can increase memory online but still need to bounce the instance to increase CPU; your Azure SQL DB stays online during scale up/down operations.
There is an alternative from Microsoft to Azure SQL DB:
“Provision a SQL Server virtual machine in Azure”
https://azure.microsoft.com/en-us/documentation/articles/virtual-machines-provision-sql-server/
A detailed explanation of the differences between the two offerings: “Understanding Azure SQL Database and SQL Server in Azure VMs”
https://azure.microsoft.com/en-us/documentation/articles/data-management-azure-sql-database-and-sql-server-iaas/
One significant difference between your stand alone SQL Server and Azure SQL DB is that with SQL DB you are paying for high levels of availability, which is achieved by running multiple instances on different machines. This would be like renting 4 of your dedicated machines and running them in an AlwaysOn Availability Group, which would change both your cost and performance. However, as you never mentioned availability, I'm guessing this isn't a concern in your scenario. SQL Server in a VM may better match your needs.
SQL DB has built in availability (which can impact performance), point in time restore capability and DR features. You have the option to scale up / down your DB based on your usage to reduce the cost. You can improve your query performance using Global query (shard data). SQl DB manages auto upgrades and patching and greatly improves the manageability story. You may need to pay a little premium for that. Application level caching / evenly distributing the load, downgrading when cold etc. may help improve your database performance and optimize the cost.
We are having significant performance problems on azure. Various factors have made this difficult to examine precisely on azure itself. If the problems are in the performance of the code or of the database I would like to examine them by running locally. However it appears that the default configuration of our database on azure is different than it is locally, e.g. apparently an azure created database defaults to run with different configuration than my local database, e.g. the default on azure includes read committed snapshot as I understand, but that is not the default for a database I create in sql server. That means that performance issues are different for the two.
My question is how can I find all such discrepancies between the setup of the two and correct them so that when I find speed issues locally I will know they represent speed issues on azure. I am a sql server novice. I recognize that I cannot recreate "time to database" and "network time" issues that way, but I don't think those are what are killing us.
You might find my answer to this post useful.
We had great advantages in implementing telemetry to gather information and use it later for analysis, to finally find out where and how you are spending your time interacting with SQL and therefore how to improve the query plans. Here is a link to the CAT blog post: http://blogs.msdn.com/b/windowsazure/archive/2013/06/28/telemetry-basics-and-troubleshooting.aspx
I read in a few places that SQL Azure data is automatically replicated and the Azure platform provides redundant copies of the data, Therefore SQL Server high availability features such as database mirroring and failover cluster aren't needed.
Has anyone got a chance to investigate deeper into this? Are all those availability enhancements really not needed in Azure? Thanks!
To clarify, I'm talking about SQL as a service and not a VM hosted SQL.
The SQL Database service (database-as-a-service) is a multi-tenant database service, and your databases are triple-replicated within the data center, providing durable storage. The service itself, being large-scale, provides high availability (since there are many VMs running the service itself, along with replicated data). Nothing is needed in terms of mirroring or failover clusters. Having said that: If, say, your particular database became unavailable for a period of time, you'll need to consider how you'll handle that situation (perhaps sync'ing to another SQL Database, maybe even in another data center).
If you go with SQL Database (DBaaS), you'll still need to work out your backup strategy, and possibly syncing with another DC (or on-premises database server) for DR purposes.
More info on SQL Database fault tolerance is here.
Your desired detail is probably contained in this MSDN article of Business Continuity and Azure SQL Database (see: http://msdn.microsoft.com/en-us/library/windowsazure/hh852669.aspx). At the most basic level Azure SQL Database will keep three replicas of your database - one primary and two secondary.
While this helps with BCP / DR scenarios you may also wish to investigate ways to backup your database so you have point-in-time restore capabilities. More information on backup / restore can be found here: http://msdn.microsoft.com/en-us/library/windowsazure/jj650016.aspx
I have been tring to use SQL Server Azure as a data storage for an Atlassian Crowd on premise for a few days and I am encountering huge performance issues.
For example, crowd admin application is extremely slow, almost unusable.
I was wondering if someone who has successfully set up this kind of solution could give me some advices.
What I have done so far :
setting up an on premise Atlassian Crowd 2.4.2 on an on premise
SQL2008R2 database
scripting the database and running the script on
an azure database (could not set up directly on azure as setup
scripts misses an azure-mandatory clustered index on table
hibernate_unique_key)
adding the mandatory clustered index to the azure hibernate_unique_key table
setting up the jdbc connection with ssl
I encounter no problem connecting crowd to the db, but everything is very slow. Crowd startup takes something like 5 minutes, when it takes something like 20 seconds with an on premise sql server.
Every round trip to the crowd admin web console takes something like 30 seconds.
My database is less than 1Mb in size. Queries execution summary in azure does not show any problematic query.
I forgot to mention that the SQL Azure Db is very reactive with SQL Server Manager or with an on premise .Net Web App
I tried both jtds jdbc driver and MS JDBC driver 4.0, both with data encryption. I tried both sqlDialect offered by crowd. It stays desperatly slow.
I tried setting special registry keys for Azure as stated by MS JDBC 4.0 driver (HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\KeepAliveTime , KeepAliveInterval, TcpMaxDataRetransmission)
Maybe it comes from :
the fact that I don't start from a clean setup done by crowd on Azure (because of the clustered index problem)
Sql Azure using utc time, making "something" expire everytime.
I would be glad if someone had advices on this problem.
Sorry - I don't have direct experience with Crowd.
I may go on a limb here, but 9 times out of 10, client applications installed remotely fail basic performance tests against SQL Database (that's the way it's called nowadays) when the application layer is exceedingly chatty (or exceedlingly chunky), with dozens or worse roundtrips for every screen/function, or returning all the records all the time. The reason this usually comes into play is due to the fact that SQL Database is far away going through a network link that is usually slower than on a local network, and on top of it the traffic is always encrypted (meaning there are more packets to transfer).
The only way to go around this type of issue, other than rewriting applications with a better design, would be to try to deploy your Crowd console in a VM in the cloud, in the same data center than you SQL Database instance. At that point, your console will be on the same network than you database, and it should, if my theory holds, be much faster.
I am working on a project which uses a relational database (SQL Server 2008). The local (on-premises) application both reads and writes to the database. I am working on a different front end for Azure (MVC2 Web Role), which will use the same data, but in a read only fashion. If I was deploying a traditional web app, I would use SQL Express to act as the local database, and deploy changes with updates to the application (the data changes very slowly) or via some sync system.
With Azure, the picture is a little cloudy (sorry, I had to). I can't seem to find any information to indicate if SQL Express will work inside of Web Roles, and if so, how to do it. Does anyone know if using SQL Express in an Azure web role is possible?
Other options I could do if forced: SQL CE or use SQL Azure. Both have a number of downsides, and are definitely less than perfect.
Thanks,
Erick
Edit
I think my scenario may not have been clear enough.
This data won't change between deployments, and is only accessed from within the Web Role; it is basically a static cache. The on-premises part is kind of a red herring, as it doesn't impact the data on the web role (aside from being its source). Basically, what I want to do is have a local data store/cache that I use existing T-SQL/DAL code with.
While I could use SQL Azure, it doesn't add anything, and if anything only adds additional overhead and failure points. I could also use a VM Role, but that is way too costly/complex.
In a perfect world, I would package the MDF into the cspkg (so it gets deployed with the app) and then use it locally from within the role. If there is no way to do this, then that is ok and I need to figure out the pros and cons of other solutions. We don't live in a perfect world. :)
You might be able to run SQL Express using a custom VHD but you won't be able to rely on any data every being present on that VHD. The VMs are completely reset when they reboot - there is no physical persistence across reboots.
If you wanted to, you might be able to locate your entire SQL Server installation in Azure blob storage.
However, in doing all of this, you'll only be able to have one worker/web role that can use that database. Remember: a SQL Server database can only be attached to one SQL Server at a time. If you want to scale out, you'll have to create new SQL Server instances for every web/worker role.
Outside of cost concerns, I can't think of anything that is in SQL Express that should be a show stopper for 99.9% of applications out there.
Adding to Jeremiah's answer: SQL Azure should give you nearly everything SQL Express does today, and you can use the Sync service to synchronize on-premise SQL Server with SQL Azure.
If you installed SQL Express into a VM role, you'd be consuming around $90 monthly just for that instance, plus blob storage (you'd want a Cloud Drive for durability). By definition, a VM Role (or any role) must support scale-out; if you were to scale to 2 instances for whatever reason, both instances would need their own copy of the database, so you'd need to create a blob snapshot for each instance.
Keep in mind, though, if you choose to install SQL Express in a VM: once you're at 2 instances, along with, say, 20GB per instance of blob storage, you're nearing $200 monthly and you're maintaining your VM's OS patches, SQL Express configuration and updates, failure recovery procedures, etc. In contrast, SQL Azure at 20GB, while costing the same $200, will offer better performance and works with the sync service, while completely removing any OS or database server management tasks from you.
To add to the already existing answers and for anyone wondering if its a good idea to run SQL Express in the cloud:
it does makes sense as a temporary storage area. Consider this architectural approach:
say you're spinning up nodes to run jobs. Storing a gazillion of calculation results might be a good idea inside a local SQL Express for each node, and provide the aggregated responses immediately when the job finishes on the node. Transfer of the no longer hot results to off-prem SQL server for future reporting/etc can be done afterwords. SQL Azure may not be optimal from the volume/latency/cost perspective to store gazillion of results and ATS will not always fit the bill, especially when relational data, performance or existing code are involved.
To expand on what David mentioned you can register for SQL Azure Data Sync CTP2 that would allow sync from SQL Server to SQL Azure here: http://www.microsoft.com/en-us/SQLAzure/datasync.aspx
Make sure to use CTP2 though since CTP1 did not support SQL Server.
If it's a read only local cache - SQL CE 4 or SQLite.
Both have Entity Framework providers.
If you're writing to it - SQL Azure