The Data Center with Killer Support

I have used Data Center’s ever since the concept of dedicated servers has existed.  A good data center is the key to a successful shared hosting service, ultimately your reputation depends on their ability to respond to problems when they occur.

I have always wanted more from a data center than just good equipment and good service.  Why could they not proactively respond to service alerts?  It would make sense as the kind of support personnel you need are right there in their office 24/7.

Some data center’s offer better service for an additional fee but this is no real improvement when you consider that you still have to manage the service.  Sure you get immediate phone access, sure they provide “personal” attention but you still need to spend time working with them to get the problem fixed.

Proactive Support

Proactive support is the business of monitoring the services provided by your server and resolving situations either before or as soon as possible without the need for customer intervention.

For example, if the load on a server begins to threaten a service a member of support is notified and then takes the necessary action to solve it.  As a dedicated server customers this is all I need in a neat package.

And there is only one company that offers this kind of support:  Liquidweb

If anyone knows any others then please tell me their story.

An example of Liquidweb’s proactive support received this morning:

Hello,

This is a Liquid Web Monitoring Notification only, and requires no response
from you. Sonar Server Health Statistics and Monitoring Services are fully
explained here: http://liquidweb.com/services/sonar.html. If you would like
more information, please reply requesting it. If you would prefer to stop
receiving these notifications, please reply with “Disable”.

For information about this and other services on your server, visit the Liquid
Web Knowledge base: http://kb.liquidweb.com/

Our monitoring service detected that named was unresponsive on your server. We
were able to log in to your server and get the service responsive again.

The following actions were taken to prevent this from happening again:

named unresponsive due to servers load, found much of the load to be from
mysql and mysql was set to the default, blank, configuration. Added the
following configuration based upon the servers memory.

[mysqld]
max_connections = 250
safe-show-database
skip-locking
key_buffer = 128M
max_allowed_packet = 1M
table_cache = 256
sort_buffer_size = 2M
read_buffer_size = 2M
read_rnd_buffer_size = 2M
myisam_sort_buffer_size = 64M
query_cache_size= 32M
max_heap_table_size = 48M
tmp_table_size = 48M
thread_cache_size = 8
wait_timeout = 30
interactive_timeout = 300

[mysqldump]
quick
max_allowed_packet = 16M

[mysql]
no-auto-rehash

[isamchk]
key_buffer = 96M
sort_buffer_size = 64M
read_buffer = 2M
write_buffer = 2M

[myisamchk]
key_buffer = 96M
sort_buffer_size = 64M
read_buffer = 2M
write_buffer = 2M

Regards,
John VanBuskirk
Monitoring Team
Systems Administrator

Tags: Hosting, Liquidweb