Forum for High Availability http://forums.mysql.com/list.php?144 Sat, 25 May 2013 04:29:50 +0000 Phorum 5.2.19 http://forums.mysql.com/read.php?144,582432,582432#msg-582432 http://forums.mysql.com/read.php?144,582432,582432#msg-582432
I don't know if this is the right place to put this question but I guess this is
related to high availablity of data.

The problem is that I have a OLTP mysql database which has two very large tables among many.
The user is interested to get a report which is created from joining this two tables. These tables are getting bigger over time with rate of millions of records
per week.

Querying the tables through on-demand joins as a response to the user's request is
very unhealthy becasuse the tables are very big and the whole database hangs during the report.

I thought one solution would be to create a second database beside the first database. The second database has a view of joint tables of the first database which is filled with the source data from the first database in an interval. sounds like staging ....

I don't know what is a standard or best practice way of doing this...
What would you suggest?
I have also familiar with most of big data definitions and solutions such as partitioning, data warehouse, staging, data marts etc. but not having had a practical experience so far, I'm not that confidence to pick a solution ...


Your suggestions are appreciated.
Thanks,
Vincent.]]> Vincent R&D High Availability Fri, 29 Mar 2013 05:42:49 +0000 http://forums.mysql.com/read.php?144,582390,582390#msg-582390 http://forums.mysql.com/read.php?144,582390,582390#msg-582390
Course Dates:

- April 11th @ 10 am PDT/1 pm EDT, https://www1.gotomeeting.com/register/821737497
- April 16th @ 10 am EDT/15:00 BST/16:00 CEST, https://www1.gotomeeting.com/register/175986064

Course Topics:

- Review of multi-master replication topologies
- Setting up Tungsten clusters for local site HA
- Linking clusters across sites
- Security for cross-site replication
- Handling local as well as site-wide failures
- Off-the-shelf filters to improvement multi-master replication
- Custom filter development to implement business rules for application messaging]]> Petri Virsunen High Availability Tue, 26 Mar 2013 18:40:19 +0000 http://forums.mysql.com/read.php?144,581393,581393#msg-581393 http://forums.mysql.com/read.php?144,581393,581393#msg-581393
my case:
My production database has around 100 tables in a single database/instance (no clusterred yet); 3 of tables are growiing very fast, in less than 1 month can use more than 1TB hard disk space, and we have to keep 6 months' data (so in a half year, disk space will be used up 6-10 TB by these 3 tables; every day there are lots of operations on these tables, update, insert and query (no delete, every operation invoked heavy API stored precedure.

questions.
1. if applying cluster, installing in 2 nodes, 5 nodes, 10 nodes, can I expect linear performance increasing to 15%,18%,20%?

2. install nodes in one physical box (usiing vmware) better or 1 node 1 box better, in terms of performance?

3. Should the storage and database seperate to get higher performance?

--
Thanks.
John]]> John Woo High Availability Thu, 21 Mar 2013 01:10:31 +0000 http://forums.mysql.com/read.php?144,579456,579456#msg-579456 http://forums.mysql.com/read.php?144,579456,579456#msg-579456
I always heard Oracle RAC is a super robust solution when it comes to availability, but how does it (2 node Oracle RAC) compare to MySQL Master-Slave (InnoDB) configuration?

They say MySQL Master/Slave can failover in about 5 to 10 seconds (automatically), and no data is lost. So the application would drop some connections while the failover takes plave (5 to 10 seconds).

What about RAC, is it a zero downtime solution? Would it keep the connections running even if the client was connected to the "faulty" server?

In summary, what makes Oracle RAC (2 nodes) so much better than MySQL Master/Slave solutions (regarding Availability only)?

Cheers]]> Jean Baro High Availability Fri, 22 Feb 2013 07:08:07 +0000 http://forums.mysql.com/read.php?144,578623,578623#msg-578623 http://forums.mysql.com/read.php?144,578623,578623#msg-578623
Course Topics
- Deployment options and techniques
- Configuring users with regard to data services
- Application configuration samples
- Read/write splitting options and demos
- Load balancing strategies
- Optimization of connection and request speed
- Understanding and fixing connectivity issues

Course Dates
- Thursday, Feb 21 @ 10 am PT/ 1 pm ET (USA & Canada)
- Tuesday, Feb 26 @ 10 am ET/ 15:00 GMT/16:00 CET (Europe)

Reserve your seat at http://bit.ly/YaP3rl]]> Petri Virsunen High Availability Wed, 06 Feb 2013 22:22:39 +0000 http://forums.mysql.com/read.php?144,576611,576611#msg-576611 http://forums.mysql.com/read.php?144,576611,576611#msg-576611
We will discuss the benefits of leveraging Continuent Tungsten clustering with MySQL, and walk you through the steps to implement a Tungsten cluster in Amazon EC2. We'll cover the prerequisites, installing and configuring Tungsten, and best practices that are part of most production installations and proof-of-concepts.


Course Topics

- Configuring MySQL and the OS for proper installation Installing a cross-site cluster
- Schema upgrade on the master database server with minimal application downtime (switch operation)
- Automated fail over when a MySQL database server crashes
- Recovery of a failed master to a fully operational slave with a single command (recover operation)
- Switching database operations to a remote site (geo-clustering, cross-site 'switch' operation)

We will also discuss and demonstrate basic operations, such as adding and removing a cluster node, basic monitoring and troubleshooting, and discuss the basic failure scenarios.

Join us to learn how to quickly configure and provision highly optimized Continuent Tungsten deployments in the cloud or on-premises! Even if you can't make the live virtual course, it is still worth registering for the course as you will receive a notification when the replay will be available, to study on-demand at your convenience.


Course Dates

- Thursday, January 17th @ 10:00 AM PST/1:00 PM EST(USA & Canada).
Reserve your seat at https://www1.gotomeeting.com/register/868426545
- Tuesday, January 22nd @ 15:00 GMT, 16:00 CET (Europe).
Reserve your seat at https://www1.gotomeeting.com/register/239773856


About Tungsten University

In the Tungsten University (http://www.continuent.com/news/live-webinars), you will learn the important details of clustering that you will need to understand to get started with Continuent Tungsten, how to properly configure and manage the cluster nodes to ensure high availability, how to install the different nodes and provide a better understanding of the internals of the cluster. Training from Tungsten University helps you keep your mission-critical applications and essential services up and running 24x7. Our live virtual courses are online instructor-led trainings that are delivered in real time via Web conferencing and audio, plus interactive enhancements such as live chat, Q&A and simulations.]]> Petri Virsunen High Availability Fri, 28 Dec 2012 18:51:42 +0000 http://forums.mysql.com/read.php?144,574777,574777#msg-574777 http://forums.mysql.com/read.php?144,574777,574777#msg-574777
121122 12:16:12 Slave SQL thread initialized, starting replication in log 'FIRST' at position 0, relay log '.\drserver-relay-bin.001' position: 4
D:\Program Files (x86)\Hornbill\Core Services\SwSqlServer\bin\SwSqlServer.exe: ready for connections.
Version: 'C.S.CB' socket: '' port: 5002
121122 12:16:12 Slave I/O thread: connected to master 'repl@master:5002', replication started in log 'FIRST' at position 4
121122 12:16:12 Master reported an unrecognized mysql version. Note that 4.0 slaves can't replicate a 5.0 or newer master.
121122 12:16:12 Slave I/O thread exiting, read up to log 'FIRST', position 4

Anyone seen this problem before? Both servers are running the same version of mysql! I'm fairly new to mysql any help is greatly appreciated.]]> James Mooney High Availability Wed, 05 Dec 2012 06:20:08 +0000 http://forums.mysql.com/read.php?144,573119,573119#msg-573119 http://forums.mysql.com/read.php?144,573119,573119#msg-573119 http://www.percona.com/webinars/industrial-strength-mysql-applications-using-percona-and-continuent

MySQL is the first choice for a large majority of web applications thanks to its ease of use and low cost of operation. However, running big apps on MySQL is still a challenge even for experts. In this webinar we will show you how to combine Percona Server and Percona XtraBackup with Continuent Tungsten to build business-critical systems capable of processing hundreds of millions of transactions per day with no downtime. We'll cover best practices from hardware configurations and MySQL tuning to HA and integration with non-MySQL systems. We will also survey technologies from Percona and Continuent that work together to help you go big on MySQL. Come learn from the foremost experts in the field of building high-performance MySQL applications!

Presenters:
- Peter Zaitsev, Founder, CEO
- Robert Hodges, CEO of Continuent

Register at http://www.percona.com/webinars/industrial-strength-mysql-applications-using-percona-and-continuent]]> Petri Virsunen High Availability Tue, 06 Nov 2012 23:13:00 +0000 http://forums.mysql.com/read.php?144,573116,573116#msg-573116 http://forums.mysql.com/read.php?144,573116,573116#msg-573116
I apologize if this is a bit of a newbie question. I've just finished setting up my first MySQL HA solution and everything is working great. I have redundant management, SQL and data nodes.

My question is this....if I point my web application to one of the SQL nodes, then I obviously lose HA. I've been doing some reading and it seems I can use mysql-proxy to point my web app to one IP which then in turn talks to both SQL nodes. However, what if the machine running mysql-proxy fails? Is there a way to make that redundant as well?

Thank you!
Jason]]> Jason Harper High Availability Tue, 06 Nov 2012 22:32:19 +0000 http://forums.mysql.com/read.php?144,572443,572443#msg-572443 http://forums.mysql.com/read.php?144,572443,572443#msg-572443 https://www1.gotomeeting.com/register/382756969]]> Petri Virsunen High Availability Mon, 29 Oct 2012 21:19:18 +0000 http://forums.mysql.com/read.php?144,571990,571990#msg-571990 http://forums.mysql.com/read.php?144,571990,571990#msg-571990
I need help configuring MySQL for special datadir and multiple mysqld instances on same host.

My client has very high PCI DSS security concerns. We are building 2 secure CentOS hosts with MySQL databases.

A is the primary host.

B is secondary.

Both have special LUKS encrypted filesystems for the actual databases.

A should replicate to B. B will become primary when A fails.

We also want B to replicate to another MySQL on another encrypted filesystem. We think that we can break replication on "C" for backup purposes, then resync afterward. One of our questions is: on which host should this 3rd database reside?

More to the point, however, is configuring MySQL for "special" datadir. The default MySQL RPM for CentOs is: mysql.x86_64 5.1.61-4.el6

Apparently, it "prefers" that datadir reside here: /var/lib/mysql

By doing this:
mysql_install_db --user=mysql --ldata=/opt/mysql01

I can then get it running by doing this:
mysqld_safe --datadir=/opt/mysql01 --user=mysql

However, /etc/rc.d/init.d/mysqld start FAILS every time

I prefer to use a fully functional init script for this, because somebody has already thought through error handling.

Please, hit me with the clue stick and guide me to best practices for this setup.

Thank you

~Mike]]> Mike Schleif High Availability Wed, 31 Oct 2012 14:31:38 +0000 http://forums.mysql.com/read.php?144,569940,569940#msg-569940 http://forums.mysql.com/read.php?144,569940,569940#msg-569940 https://www1.gotomeeting.com/register/152529257


Cross-site databases are the next challenge facing today's MySQL-based businesses. Continuent Tungsten provides multiple options for spreading data across sites, including primary/DR, multi-master, and system-of-record approaches. Join us to learn how Continuent Tungsten enables replication, failover, and routing of transactions between sites. We cover the following topics:

- What are the differences between using disaster recovery sites and full multi-master?
- What are some of the key things applications must do in order to prepare for multi-master operation?
- What is system-of-record and how does it work?
- How do you set up multiple site data management with Continuent Tungsten?
- How do you fail over between sites?
- How does Continuent Tungsten route transactions transparently between sites?

Continuent is leading the charge to make multi-site operations work for SQL databases. Learn how Continuent Tungsten together with MySQL can give you the advanced data management capabilities your business needs today!]]> Petri Virsunen High Availability Mon, 01 Oct 2012 18:55:18 +0000 http://forums.mysql.com/read.php?144,569736,569736#msg-569736 http://forums.mysql.com/read.php?144,569736,569736#msg-569736
I am given a task to identify best Fault Tolerant HA topology with 2 MySQL Servers.

Little background:-

1. Our application at present runs over Oracle Database. We are moving to MySQL for Lower cost Solution for Emerging Telecom Markets
2. Performance is not too much of a concern.
3. Dataset and user base is low.
4. It is just that solution should be robust when it comes to Fault tolerance and HA.

I have researched for a few days and have following topologies in mind:-


1. MySQL Multi-Master replication with Heartbeat and Pacemaker (No DRBD): For such a topology we would have one Floating IP Plumbed on any one node.

In case of a node/service/network/disk failure this floating IP will shift to the other active node (and send and alarm mail) and process request with minimal service downtime.

Advantages:-

a) Slave Promotion not required.
b) Single IP HA connection
c) Other master can still be used for DB backups etc.

Disadvantages:-

a) Read-only queries cannot be offloaded to the other master.
b) High speed Network connectivity required for heartbeat
c) Support fee (LINBIT for Heartbeat and Pacemaker).
d) We need some sort of Replication monitoring scripts which will raise alarm in case replication is not working or in case both the nodes are out of sync.


2. MySQL Multi-master replication with ReplicationDriver using J Connector (No Heartbeat/Pacemaker): Our application will always make DB connections via JDBC drivers.

There is no floating IP in this case. We will be using connector replication driver to manage our connections. In this case we will tell replication driver to use one master as master and the other as a slave (so that we do not run into Update Conflicts).

Advantages:-

a) Slave Promotion not required.
b) Other master can be used for DB backups and read only queries etc.
c) No LINBIT support fee
d) No network interconnect required for heartbeat
e) less complex to implement

Disadvantages:-

a) No Single IP - DB access code should be intelligent so as to write only to one master at a time.
b) We need some sort of Replication monitoring scripts which will raise alarm in case replication is not working or in case both the nodes are out of sync.

Similarly we have following configurations


3. MySQL with DRBD, heartbeat and pacemaker (No Replication): In this case MySQL will be running on one node at a time and DRBD will keep the two nodes in Sync.

Advantages:-

a) Synchronous – MySQL replication can at most be semi-synchronous
b) No replication monitoring scripts required.

Disadvantages:-

a) LINBIT Support Fee
b) One node is always down – no offloading possible.
c) More Failover time as MySQL is required to be brought up.
d) High Complexity.


I also considered 2 more topologies – Same as 1 & 2 but instead of Multi-Master it will be master-slave. This would have an overhead of Slave Promotion and Master Conversion to slave.

Please advice which is the best HA topology to have. Also I have no idea about stability of each of these topologies as my knowledge is primarily theoretical.

Thank you.

Best Regards,
Prashant]]> Prashant Goel High Availability Fri, 05 Oct 2012 04:10:42 +0000 http://forums.mysql.com/read.php?144,569012,569012#msg-569012 http://forums.mysql.com/read.php?144,569012,569012#msg-569012 https://www1.gotomeeting.com/register/575821320

High availability is about more than making sure that apps can get to your data even if there is a failure. How about when you are upgrading your database schema? What if you need to add memory to a database server or reconfigure/restart MySQL? If your apps want to read data from a MySQL slave, how can you be sure they are not reading stale data without re-coding your apps?
What if your main data center takes a hit? And what if you want better MySQL stability, performance, and simplified administration, but you cannot or do not want to re-write your apps or convert your current data to run with a proprietary MySQL re-write or a promised new version of MySQL?

During this webinar, we will go through a detailed case study to show you how you can start with the MySQL version/build (MySQL Community 5.x, MySQL Enterprise 5.x, MariaDB, Percona, etc.), along with the data and apps that you already have, and iteratively build up a highly-available, scalable, multi-site database infrastructure where you can perform complex data management tasks with single commands.

You will see live demos of the following:
- Transition from MySQL 'native' replication to a Continuent Tungsten cluster with zero app downtime
- Reading consistent data from slaves without app changes (SmartScale)
- Schema upgrade on the master database server with no app downtime (switch operation)
- Automated fail over when a MySQL database server crashes
- Recovery of a failed master to a fully operational slave with a single command
- Switch of database operations to a remote site with no app down time (geo-clustering, cross-site 'switch' operation).

Join us at https://www1.gotomeeting.com/register/575821320 !]]> Petri Virsunen High Availability Wed, 19 Sep 2012 20:16:26 +0000 http://forums.mysql.com/read.php?144,566803,566803#msg-566803 http://forums.mysql.com/read.php?144,566803,566803#msg-566803 Steve McGarrah High Availability Thu, 30 Aug 2012 18:16:36 +0000 http://forums.mysql.com/read.php?144,561888,561888#msg-561888 http://forums.mysql.com/read.php?144,561888,561888#msg-561888
Backgrounder:
- We using mysql on Ubuntu servers using heartbeat/pacemaker in an Active/Passive redundant pair configuration.
- When simulating failover, we have observed that "service mysql start" on the new server will exit with success although mysql has begun the recovery processing and "server mysql status" reports that mysql is not available.
- so we need prevent heartbeat/pacemaker from taking any action while mysql is doing its recovery processing

is there a way to determine that mysql is doing the recovery processing?]]> Mike Denny High Availability Tue, 28 Aug 2012 14:47:38 +0000 http://forums.mysql.com/read.php?144,561670,561670#msg-561670 http://forums.mysql.com/read.php?144,561670,561670#msg-561670
I installed PHP and MySQL,

Put a link PHP Mysql web application that works windows server 2008R2 IIS7, SELECT, UPDATE, DELETE, but does not function INSERT.

The same application other server Xampp is running at 100%.


Ernest]]> ernest francisc High Availability Sun, 22 Jul 2012 04:54:24 +0000 http://forums.mysql.com/read.php?144,560522,560522#msg-560522 http://forums.mysql.com/read.php?144,560522,560522#msg-560522 Everything seems to work fine and the performance seems to meet our expectations. However, I notice during our testing that after I run mysql under a load (sysbench) for a while, the shutdown times are in terms of hours? I have reducd the innodb log size to 32M and it still takes hours. A similar test using local disks takes about 5 minutes to shutdown.

Has anyone seen this great of disparity and long shutdown times when using NFS?
Any guidance? Although I have not tested it, I suspect that the recovery times would be about the as the shutdown times.]]> Gil McGrath High Availability Sat, 14 Jul 2012 14:23:34 +0000 http://forums.mysql.com/read.php?144,558753,558753#msg-558753 http://forums.mysql.com/read.php?144,558753,558753#msg-558753
Is this expected? Is it possible for MySQL to support a hard, real-time application without occasional excessive insert times?

Many thanks!]]> Todd Sampson High Availability Fri, 29 Jun 2012 14:55:26 +0000 http://forums.mysql.com/read.php?144,558312,558312#msg-558312 http://forums.mysql.com/read.php?144,558312,558312#msg-558312
I want to setup MySQL in a Windows 2008 R2 failover cluster but I face to a little problem...
The cluster uses 2 nodes and MySql 5.5 is installed as a service on both. The datadir is set on a cluster shared volume (CSV) to be accessible from both nodes.

I added MySQL service as a cluster's resources, called "MYSQL", set with a private IP (10.10.10.10) declare as static in a WINS server (I've no access to my company DNS). The MySql services on each node are in manual mode and only one is started (on node1). MySql works fine, I can set new databases and work on them.

For testing the failover function, I set a php website that uses MySql database. If MySql service is started on node1, all goes fine. But if I simulate the fail of node1, the MySql resource (and service) starts on node2 but the website can't get connexion to the database anymore.

After many hours working on this problem, I'm unable to fix it. If someone can give me some advice, it'll be great !
Thx in advance !]]> Eric DUCAS High Availability Fri, 22 Jun 2012 14:12:58 +0000 http://forums.mysql.com/read.php?144,558049,558049#msg-558049 http://forums.mysql.com/read.php?144,558049,558049#msg-558049
I have several days trying to solve this problem,

The thing is I'm looking for a way to redirect connected clients from one server to another.



For example, I have 10 connections (some limit set) in a mysql server, the eleven client reaches, I try to do is redirect that client, either before or after you connect to the server for a backup


Customers are programs written in C that are in constant communication with the database.

I wish someone could help me with this problem

thank you very much!]]> Edgar Juve High Availability Sat, 30 Jun 2012 00:28:41 +0000 http://forums.mysql.com/read.php?144,551832,551832#msg-551832 http://forums.mysql.com/read.php?144,551832,551832#msg-551832 http://www.slideshare.net/nixnutz/load-mysq-clusterin-balancing-peclmysqlndms-14
http://www.slideshare.net/nixnutz/the-mysqlnd-replication-and-load-balancing-plugin
http://blog.ulf-wendel.de/2012/peclmysqlnd_ms-14-load-mysql-clustering-balancing/
http://blog.ulf-wendel.de/2011/the-mysqlnd-replication-plugin-110-release/
http://blog.ulf-wendel.de/2011/replication-plugin-filter-and-conquer/
http://blog.ulf-wendel.de/2011/peclmysqlnd_ms-compared-to-a-classic/]]> Edwin Desouza High Availability Wed, 06 Jun 2012 17:11:50 +0000 http://forums.mysql.com/read.php?144,550403,550403#msg-550403 http://forums.mysql.com/read.php?144,550403,550403#msg-550403
Join Heather Sullivan, Senior Database Administrator at Constant Contact at https://www1.gotomeeting.com/register/581656488 to learn why Constant Contact chose Continuent’s Tungsten for their replication needs. Heather discusses how Tungsten provides a means to manage a highly available environment with disaster recovery, and greatly simplifies the day-to-day management of MySQL replication. Continuent Tungsten also helps to lay the foundation for Constant Contact to expand its infrastructure footprint to best serve customers.]]> Petri Virsunen High Availability Tue, 05 Jun 2012 21:09:35 +0000 http://forums.mysql.com/read.php?144,532647,532647#msg-532647 http://forums.mysql.com/read.php?144,532647,532647#msg-532647 https://www1.gotomeeting.com/register/733210944. Learn how you can start with the data and apps you already have, and build a highly-available, scalable multi-site infrastructure. Thursday 5/17 @ 1 pm ET/10 am PT]]> Petri Virsunen High Availability Wed, 09 May 2012 20:21:44 +0000 http://forums.mysql.com/read.php?144,527547,527547#msg-527547 http://forums.mysql.com/read.php?144,527547,527547#msg-527547 https://www1.gotomeeting.com/register/656585273

Continuent Tungsten has broken new ground to make previously complex or completely unattainable MySQL data management solutions simple and easy to deploy. Continuent Tungsten brings high availability, performance scaling, and simple cluster management to off-the-shelf MySQL. Continuent Tungsten also includes advanced multi-master topologies that enable users to spread data across clusters and sites.

In this webinar we show how to architect and manage cutting edge data management topologies. We start with simple clustering, build up to multi-master/multi-site, and end by illustrating how to integrate your MySQL systems to other DBMS types. The webinar includes live demos of simple failover, DR site setup, and fan-in from multiple sites to a single reporting server. By the end of the talk you will understand how Continuent Tungsten works and the problems you can solve with it.

Continuent Tungsten customers include American Express, AT&T, Constant Contact, F-Secure, Financial Times, Juniper Networks, Hearst Magazines UK, Marketo, Norwegian Broadcasting Corporation, Sony Music Entertainment, Thomson Reuters and Zappos.com, an Amazon company. The largest Tungsten Enterprise deployment processes 500,000,000 transactions per day on 150TB of managed data.

Join us at https://www1.gotomeeting.com/register/656585273 to learn how you can put the power Tungsten to use today!]]> Petri Virsunen High Availability Wed, 25 Apr 2012 20:20:45 +0000 http://forums.mysql.com/read.php?144,526339,526339#msg-526339 http://forums.mysql.com/read.php?144,526339,526339#msg-526339 Chris West High Availability Sat, 21 Apr 2012 05:34:16 +0000 http://forums.mysql.com/read.php?144,522462,522462#msg-522462 http://forums.mysql.com/read.php?144,522462,522462#msg-522462
I am working at a University on a high-profile database driven project that we expect to be slammed within the first few months. Since this is a new project and one that we expect to be popular, we don't know what kind of usage to expect, but we want to be prepared. Therefore, we are building in extra capacity.

Our top goals are scalability and high availability, provided we hope through multiple MySQL nodes and VMWare functionality. I've been surprised that there are not more MySQL architects trying to meet these high-level goals using virtualization and shared storage (or at least they do not seem to be writing about it).

I've looked at replication, multi-mastering, DRBD, clustering, partitioning, and sharding.

Here's what we got, and some of our constraints:

* We are concerned that One Big Database instance won't be enough to handle all of the queries, plus it is a single point of failure. Therefore, multiple nodes are desirable.

* With the primary application that will be using the database, writes and reads cannot be split off from each other. This limitation alone, rules out replication, MMM, and a few other solutions.

* We do not expect to be especially write-heavy.

* We have shared storage in the form of an iSCSI SAN. We'd like to leverage the shared storage, if possible.

* We have VMWare HA which already monitors hosts and brings them up within minutes elsewhere if we lose a host. So some of the suggested HA solutions are redundant.

* We expect to have another instance of our system running in the Amazon cloud for the first few months while the traffic is high, so we may take advantage of RDS, though an exact duplicate of our local system will save us development work.

Thanks for any advice you can give.

Wes Modes]]> Wes Modes High Availability Fri, 30 Mar 2012 01:18:57 +0000 http://forums.mysql.com/read.php?144,521510,521510#msg-521510 http://forums.mysql.com/read.php?144,521510,521510#msg-521510 https://www1.gotomeeting.com/register/429957080

Tuesday, 4/3 @ 9 am ET/15:00 CEST/14:00 BST/16:00 EEST
https://www1.gotomeeting.com/register/118613520

Your company has built a great new app and launched it in the cloud. And now you are seeing what many wish for: an exponential adoption of your app. But is your database tier really up to the job?

- What happens if your MySQL server fails?
- Can you fail over to a replica quickly?
- Do you even have a replica to fail over to?
- Can you keep your app available even as you upgrade your database schema or add additional or more powerful database servers?
- Can you easily scale read operations by load balancing them across a set of replicas?
- Do you have a way to maintain a set of data that is globally accessible, across Amazon EC2 regions in the USA, Europe, APAC, Japan etc., and that allows app users to get to and update their data with minimal latency, no matter where they are located?
- How do you make sure that an update made in a database in Japan gets propagated to databases in Europe, USA, APAC etc. with minimal latency?

This webinar focuses on the ways that Tungsten Enterprise can help you to solve these issues by combining local MySQL clusters, for high-availability and scalability, with cross-region, multi-master replication to maintain a globally distributed database.]]> Petri Virsunen High Availability Fri, 23 Mar 2012 22:32:49 +0000 http://forums.mysql.com/read.php?144,519236,519236#msg-519236 http://forums.mysql.com/read.php?144,519236,519236#msg-519236
I'm currently building up a HA Cluster using lastest MySQL Cluster, Heartbeat and Ha-Proxy.

Basically I already have setup everything needed an it seems to work fine. Now I've stumbled across a big problem. I dumped one database from our production server to do some failover testing on the cluster. The problem is that in order to have the db replicated on both storage nodes I have to change ENGINE in the sqldump to NDB. Once I did that and tried to restore the db I got errors saying that NDB is not capable of e.g. FULLTEXT searches etc. Googling the web shows clearly that I have to use MyISAM since NDB doesn't support it (yet???). Now I'm stuck becuase if I can't replicate a db my load-balancer (HA-Proxy) is useless since I don't have the db on both storage nodes. Once a client is switched to the other storage node there are no tables in the db. :( I read a few posts about using sphinx when myISAM is using Fulltext searches. Can I simply change some lines in the sql dump so it uses Sphinx?

Can I use on storage node with two mysqld's? The storage nodes are conntected via iSCSI to a netapp storage that is clustered so I would not really need a second storage node.]]> Oliver Weinmann High Availability Sat, 21 Apr 2012 19:32:14 +0000 http://forums.mysql.com/read.php?144,517131,517131#msg-517131 http://forums.mysql.com/read.php?144,517131,517131#msg-517131 https://www1.gotomeeting.com/register/597680512, and
Tuesday 3/6/12 @ 15:00 CET/14:00 GMT
https://www1.gotomeeting.com/register/860877873

Learn how the next version of Tungsten Enterprise will help you push the envelope in multi-master, multi-site MySQL database clusters. We will be introducing the following features:

- Setting up MySQL clusters with a single installation command
- Automatic master failover without data loss
- Building disaster recovery sites quickly
- Automated failover between sites
- Building multi-master MySQL clusters that cross multiple worldwide sites, in the cloud.

Tungsten Enterprise is setting the standard for helping business build highly available and fast database systems using any MySQL version (MySQL Community 5.x, MySQL Enterprise 5.x, MariaDB, XtraDB). Our customers include Constant Contact, F-Secure, Hearst Magazines UK, Marketo, Norwegian Broadcasting Corporation, Sony Music Entertainment, Thomson Reuters and Zappos.com.

Join us to learn how Tungsten Enterprise 1.5 helps you in managing your data worldwide! Register at http://www.continuent.com/news/webinars!]]> Petri Virsunen High Availability Fri, 24 Feb 2012 17:18:04 +0000