8 posts tagged “ec2”

I did read about this a while back when the challenges using EC2 was how not lose a database in case of a machine failure. Luckily it never happened and forgot all about EBS, the persistent disk for Amazon EC2. We went live with a new customer over the weekend and EBS came up all over again for data backup. Quite simple to use it.

Interesting aspect to tackle after putting applications on EC2 in a live production environment. Our first customer has been live for over 1 year on EC2 and never been down for a single time. We didn't have to reboot the linux server though we had to restart our java processes for upgrades. Touchwood. Database incremental backups happen every minute to S3 and a daily back up at night. This is mostly an enterprise app, so not very stringent requirements on real time failover. Shortly we are going live with a new retail customer facing application - a 24 x 7 uptime requirement. Scalability is not much of an issue as our server is still operating under 65MB with response times under 50ms most of the time. Failover is a key aspect for us - passive or active, active being the preferred option.

Ant (http://ant/.apache.org) is embodies simplicity, extensibility and never in your way software. A lot of developer tasks can be automated with this tool. For quite some time we used it to just build a WAR file and deploy it Tomcat on our desktop machines. Deployment to server used to be a one hour task. Build with the server configuration files, ftp to server, extract static stuff to apache, re-deploy the database and recycle the server.

Recently my EC2 instance was hung. The website was getting timed out. A clear signal that the web server is not responding and all socket communication is not happening. As the sockets were not reactive, so the ssh server. I was not able to login using the SSH terminal - PUTTY.

Now I just can't shut down that instance and start another one as I had some production data in mysql data files which had not backed up. I had told my customer 99% uptime, 256MBps connectivity but what about reliability.. Now they are going to lose production data. Literally, I was shocked. I desparetly tried to get my data back. Did googling and got depressed reading the comments of pool. We then put a urgent request in the forum and amazon replied that nothing is possible.

After I cooled down a bit, it was time to fight back. First thing I thought of trying was restarting the instance. However, I was not sure about my data. All my mysql data files were in /mnt. So the question was am I going to loose data if I restart my instance. Some links I referred confused me more; I left with a understanding that anything at MNT has the possibility of loss. But the mnt was my target. Now I took a decision to restart my server - The last resort.

I issued restart command from my desktop. Oops... there was no effect. Then I went to my another amazon instance. There I had the EC2 tools. I used those tools and issued a restart command. Did a ssh again. The server was still hung.  I sat tight for 5 mins. Again connected via my putty. Bingo, this time the server shown me the login prompt. My curiosity doubled up.. Am I going to have my data in /mnt. Then I changed directory to /mnt. Everything was there intact. I just relaxed. Thank God - I  survived a crash 

I remember few years back when I was running the Websphere, it was very difficult to kill a instance. On killing of one instance, a deamon was spawning another. Killing that one, one more was coming up. I had tried "kill -9", "kill -KILL" - No help. When one was going down, other was coming up. It is not like, I have 2 process instances running and on killing of both, the application  is completely down. Rather, it is only one and taking birth each time it is murdered.

Why not this happens for a server. If I consider that is possible, virtually the application is immortal. 

[I have taken this section out from my regular writting as I want to highlight on one particular issue. Recently I had mapped a production instance to amazon EC2 instance in web DNS mapping. I issued a shutdown command on EC2 machine by mistake. It just took me 2 mins to bring up a snapshot of the last instance as I was taking the snapshot. Then I pointed to this new server address in web DNS mapping. This process took a good 30mins and the system remained down for so long :(]

Consider a machinsm is available, where web DNS mapping is instantaneous to the depployed server. In this spectrum, I have an application which transcends through server instances. The possibilities lies on the snapshot creation. Before death can I take an exact snapshot and all next instances starts from there. Amazon EC2 has infiniband networking to S3. So, network latency is not my concern. Hadoop is there for my rescue to transfer the big files. The big file is the tar ball of the complete mnt. If you are reading my previous posts, you know you can shove everyting to mnt. That means the complete harddisk portion of my interest to the S3. This helps on saving the database files, other contents, application state and everything else.

During shutdown,

1. Create a snapshot tar ball of /mnt of the portion of harddisk of my interest. - 30sec
2. Copy this to S3.
3. Start another instance.
4. When the next instance comes up, it self configures itself during start up taking the file from S3.

I have taken snapshot and created a new instance quickly but manually. It takes just a minute. How can I do it automatically?

First I looked for linux shutdown hooks like init. No luck. If you are aware of any, please let me know. Now I am going to explore by creating a blank java application and starting it during init and putting a shutdown hook there. This shutdown hook will grap the linux shutdown. I am concluding this reading the post at http://forum.java.sun.com/thread.jspa?threadID=504889 . Once I finish this experiement, I blog about my findings :) This will open up following possibilities in EC2:

1. Immortal application in a mortal server
2. No information loss when instance shutted down.
3. Many app instance creation when load increases. [What about the database??? I have a big blog coming on this. I am half way through]

Don't forget to help me on implementing a linux hook of shutdown rather than going via the java way...  A work around not a clean solution :(

Software Architecture For Developing Countries

I was monitoring my EC2 linux instance. I suddenly found my free memory is depleting. I was cluless. With research I got an paper which explains about the linux cache memory. http://mail.nl.linux.org/linux-mm/2003-03/msg00077.html 

    Memory Cached - Any modern operating system will cache files frequently accessed.

With detail research I found out one of the major I/O operations are log files. I use log4J for logging in the application. Though the log files are rotated, the system spits out huge amount of logging information. The monitoring scripts, scans through these log files in each 5mins to get if any exception has happend. The frequent access of log files and huge log file sizes makes the OS caching these files.

Major loggers of log4J are only in WARN level. However, each request entry is written for monitoring purpose. Then I asked myself why am I writting these logs. The answer was

1. For trouble shooting during a fault
2. For Monitoring
3. As a audit mechanism

Consider system has a fault / exception. This time I require to log the complete execution path as in case of debug to quickly identify the issue and solve it. With warning, it becomes difficult to get closure to actual circumstancecs.

Regarding this I had discussed with sunil before and he was talking about a in memory logging and on need flushing. This is the background of conceiving onelineLog -

A simple on need logging framework to optimize system resources and serving clients faster.

Software Architecture For Developing Countries

Software Architecture For Developing Countries

The definition of availability:

Present and ready for use; at hand; accessible:
Capable of being gotten; obtainable
Qualified and willing to be of service or assistance

Unavailability is not felt till the moment users reach out for the service and are not served. For me the counting of unavailability starts the moment absence of availability felt.

My application is a web based application. The potential of unavailability is bacause of:

• Absence of the desktop machines.
• Absence of network connection.
• Unable to serve due to fault on:
1. Website not available
2. Application servers are down or hung
3. Slow performance and high response time
4. Software bug and functionality is not available.

To make a high available application, the system requires:

1. Mobile presence
2. Smart client applications, provision to browse at client side during no network.
3. The website is up all the time.All machines and mobile devices are functioning properly.
4. All serves are functioning properly - Apache, tomcat, mysql, mail
5. All service agents are functioning properly - sms, email

During,

1. Single node failure
2. High load situation
3. New application version deployment
4. Regular machine patch management
5. Hacking attacks, intrusion attacks
6. Virus Attack
7. External applications unavailability

and immediately serving after a fault due to,

• Database failure
• System/node unavailability
• Application bugs

OK, the principles I am guided to provide high availability in my application are:

During Requirement

Availability for me is a functional requirement. All the requirements are analyzed from this perspective too. If system couldn't be available from world wide web, can he perform some operations from his mobile device via SMS, can he send a mail and do the same operation.

During Architecture & Design

• Stateless architecture on Amazon EC2 to quickly provide more servers as the number of users increase.
• Central point of information update and numerious points of information viewing to scale. In my application around 80% time information is viewed and 20% time it get's modified.
• Each component handles failure situation. Ex. The email agent if could not send mail using account X, it will try another email provider.

During execution - production environment

On production - During available period:

1. Preparing Stand by machine to take over during failure
2. Monitoring the health and providing necessary support.
3. Continuous revision of server deployment according to the measured data: 
      Staep -1 )   Moving database and other single point of failure to an dedicated machine.
      Step - 2)  Up scaling te machine by adding more memory, cpu, disk and other resources in the existing machine
      Step - 3) Hot backup and replication of all changed content.

On production - During Unavailable period

1. Browser always only refer to one url. This is the first single point of failure. To avoid this, we have a javascript at our yahoo hosting site. Application first loads the infromation from here. This failover is provided by yahoo. Now the web application stack maintains the list of one primary and another secondary server for processing at amazon instances. In case of access failure to primary, the browser switches to the secondary servers.
2. As the apache and tomcat servers are stateless. Once the server goes down, immediately a new instance is started by the clustered server.
3. Once the master database goes down, immediate replica takes over and chooses it's own back up node. It passes the next backup to all the clients when they connect to master for sync up. Master node pushes information to all the subscribed nodes. All updates, deletes, inserts happen in master node. Child nodes serve the lookup(select) queries.
4. Once a database replica goes down or a new instance is created, it fist takes yesterday backup and then replicates today's transaction querying from master node.

The Preparedness

1. Failover drill process. It forces the system to shut down and waits for the failover system to take care.

Software Architecture For Developing Countries