In an ongoing effort to provide the highest level of network and IP services to our customers, we will hold a network maintenance window tonight night, August 8th between 1:00 AM and 3:00 AM.

The purpose of this maintenance window is to move access layer links in our C2 datacenter to more powerful distribution switches.

Customers will notice short periods of connection latency lasting no more than a few minutes and/or TCP session resets as traffic is stabilized on the new switching platform. In other words, “sites will be slow”.

Servers themselves will not be affected or brought offline. This work will related only to the switching units (network connections) that feed the servers.

Below is our monthly uptime report for June 2009. While we strive for 100% uptime, sometimes issues arise.

————————————-
www.sononaco.com
Service: Sononaco 4
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
www.spromail.com
Service: Sononaco 1
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
64.237.99.32
Service: Sononaco 3
Outages: 2 (1 for Network maintenance, 1 for hardware maintenance)
Downtime: 1 hrs, 10 mins
Uptime: 99.84%
————————————-
mail.sononaco.com
Service: Zimbra Mail Server
Outages: 1
Downtime: 0 hrs, 20 mins
Uptime: 99.95%
Downtime was documented and was due to mail server upgrades.
————————————-
[Hosted Client]
Service: [Hosted Client]
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-

Below is our monthly uptime report for May 2009. While we strive for 100% uptime, sometimes issues arise. Fortunately due to the configurations we have in place for our hosting, no data was lost.

————————————-
www.sononaco.com
Service: Sononaco 4
Outages: 1 – 5/21 at 1:20 AM for 0.064 seconds
Downtime: 0 hrs, 5 mins
Uptime: 99.99%
————————————-
www.spromail.com
Service: Sononaco 1
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
64.237.99.32
Service: Sononaco 3
Outages: 1 – 5/13 at 12:15 PM for 0.032 seconds
Downtime: 0 hrs, 5 mins
Uptime: 99.99%
————————————-
mail.sononaco.com
Service: Zimbra Mail Server
Outages: 2:
May 9, 12:40 PM for 25 seconds – Power cycling at the hosting center
May 15, 2009 at 6:32 PM – Hard drive controller board failure
Downtime: 3 hrs, 45 mins
Uptime: 99.51%
The outages were documented. Please see previous blog entries about the hard drive crash for more information on this incident.
————————————-
[Hosted Client]
Service: [Hosted Client]
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-

If even Google can go down, nobody is safe.

On Friday May 15th at approximately 6:30 PM one of the circuit boards went out on one of the three hard drives in our mail server. Our technicians worked quickly to repair the problem.

Unfortunately, the screws on the hard drive are a rare Torx type of screw. Our technicians had to track down the proper toolset to remove the hard drive.

Our mail server is running in what is referred to as a RAID5 configuration. What this means is that there are three hard drives in the server that share all of the data on the system as a single drive. We can lose two drives and not lose any data, because that data is shared across the three.

Once the hard drive was repaired the RAID array took about half an hour to rebuild and mail functions were restored around 10 PM. Any e-mail received during this time was eventually delivered over the next couple of hours.

We apologize for the downtime. With the new controller board installed on the hard drive we hope to avoid such situations in the future.

We are in the process of upgrading all of our hosted sites and codebases to Zend Framework 1.8.1. If you are a developer please be sure to update your local copy of the Framework.

Zend released version 1.8 of the Zend Framework around the first of May. Sononaco is upgrading the Zend Framework running on all of our servers. There are several notable changes to the code but there is one issue that needs to be immediately addressed: the Autoloader.

Beginning today (May 13th) we will begin upgrading the Framework on our staging/testing server, (at 66.240.223.252) and our Zend Server. All Zend Server sites have been upgraded to use the 1.8.1 bootstrapper. All site on the “dev” server include any site with the domain name beginning with “dev” or “staging”.

How does this impact you? Unless you ar a developer, not much at all. If you are one of the developers using the Zend Framework then you will need to do a couple of things:

First, update your local working copy of the site. Effective immediately all Application codebases in Subversion should contain the code for the 1.8 Framework. The 1.8.x Bootstrapper code is not compatible with versions of Zend Framework older than 1.8.x!!

If you run the new Application code against the 1.7.x Framework you will see errors so be sure to upgrade your local Framework to 1.8.x.

For the moment we are not upgrading existing bootstrapers or Application layouts to the new Zend_Application_Boostrapper. All projects going forward will contain the new Bootstrapper system.

You can download version 1.8.x of the Zend Framework here.

This is a report we receive monthly that reports on our hosting uptime. I thought this would be interesting to share each month.

————————————-

Below is your monthly uptime report for April 2009

Please do not reply to this email.
For further assistance, please contact support[at]siteuptime.com.

————————————-
www.sononaco.com
Service: Sononaco 4
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
www.spromail.com
Service: Sononaco 1
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
64.237.99.32
Service: Sononaco 3
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
mail.sononaco.com
Service: Zimbra Mail Server
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-
[Hosted Client]
Service: [Hosted Client]
Outages: 0
Downtime: 0 hrs, 0 mins
Uptime: 100.00%
————————————-

The Zimbra mail server upgrade has been completed.

As you may know, we have taken the primary web hosting server offline several times within the past week to insert additional RAM. How hard is it to upgrade RAM? We’ll explain.

Is running out of RAM on a server a bad thing? Yes and no. That means someone is doing something right and getting lots of traffic! It also needs we need more RAM.

The first time the new RAM was installed the server starting giving a BIOS error. This is essentially an error that states the RAM isn’t compatible with the server.

New RAM was acquired and we tried installing it again. Unfortunately it seems a similar error crept up. There appears to be an incompatibility issue with the motherboard.

So, where do we go next?

New motherboard.

The new motherboard will be installed in the coming days. Updates will be posted on the blog and in the news. Please keep checking back.