XILO News and Status

We’ll be conducting routine maintenance on Indigo this evening. During the maintenance period we’ll be removing it from the cluster that it is currently located on.

The reason for this is that after the issue 2 weeks ago with our cluster, we have decided to revert back to our standard configuration of hosted servers so that such an outage is avoided in the future.

We’ll be starting the work at 11pm GMT this evening and we expect to be completed by 3am. Although the window is 4 hours, we expect the work to be completed in a much shorter period.

During the period, we will be disabling the following services to prevent any differences in data during the migration;

• FTP
• SMTP
• POP3/IMAP
• SCP(SSH)

Please accept our apologies for the short notice for the work. This is due to a few residual issues that are causing higher than normal loads on Indigo.

If you have any questions, please do contact us via My XILO.

Due to a customer script, the KILO server had two unplanned and necessary reboots today. The script in question saturated the CPU and as such, we were unable to login via SSH or console.

After the reboot, we identified the customer and have suspended their account until we can resolve their issues.

Unfortunately, due to the unclean reboot - the RAID array is having to initialise the data on the array. This will take a few hours and during this period, the server will respond slightly slower than normal.

Apologies for the two small outages and the slower than normal response times whilst the RAID initialises

We’d like to remind all our customers about using secure passwords on their hosting accounts. This includes FTP, mail box and MySQL (database) accounts. Also, for our resellers, creating accounts for your clients.

We’re seeing more and more clients and resellers using simple passwords similar to the below;

• 123456
• test
• password1
• testtest
• 123

This can lead to serious security issues on your account which may lead to the suspension of your account(s) if the password is guessed or brute forced.

We’ll be activating a new feature in cPanel that will only allow the creation of slightly harder password combinations to prevent any accounts being accessed without authorisation.  This is in addition to the brute force detection that we currently have in place which blocks most attacks.

If you feel you are having problems changing your password, please contact support via my.xilo.net.

We’ll be changing the way some of our older reseller accounts were assigned nameservers and IP addressing.

This will affect any reseller that has private nameservers (such as ns1.mydomain.co.uk etc) or is using any of the following;

• ns1.alpha.*.net (and ns2)
• ns1.bravo.*.net (and ns2)
• ns1.charlie.*.net (and ns2)
• ns1.delta.*.net (and ns2)
• ns1.echo.*.net (and ns2)
• ns1.foxtrot.*.net (and ns2)
• ns1.kilo.*.net (and ns2)
• ns1.india.*.net (and ns2)

(* - we have removed the domain to protect our resellers)

Also, any customers that are using the following IP for any of their private nameservers will need to update to the new, permanent value.

• Old : 72.21.47.31
• New: 67.225.140.31

This should be changed in your zone via WHM as well as at your domain registry. If you are unsure on how to do either of these, please contact support via http://my.xilo.net/ and we’ll take care of it for you.

The old nameservers will cease to work after 1st March 2008 so it is imperative that you change these immediately. Please contact support for the new nameservers for all of your accounts if you do not have private nameservers assigned. (Once again, not posted for the protection of resellers)

If you have any questions, please do contact us.

We’re currently rebuilding the cluster platform for the servers above.

Unfortuantely, after the issues relating to power, it has severely corrupted the storage array. We have attempted to restore from our back cluster but this has failed. We have contacted our backup software provider to ask them for assistance.

We’re also currently storing the data from the damaged array to another server so we can at least start a manual rebuild.

Please bookmark this page if possible instead of raising tickets or phone calls as we’re extremely busy and not able to answer all calls in a timely manner.

We’ll update this as much as possible and hope to have service restored soon.

At around 3am, BlueSquare 1 experienced cooling issues which resulted in the overheating of several servers in the facility.

Unfortunately, due to an apparent knock-on effect, this tripped a power bar in one of our racks which resulted in the fuse blowing causing half of the PDU (power distribution unit) to stop supplying power to some servers and equipment.

Engineers are onsite and are trying to get back on all servers as soon as possible. Some have blown PSU (power supply units) because of this which is delaying matters.

A full outage report will be sent as soon as possible.

Following our PHP5 upgrades on our reseller hosting platform, we will now be upgrading our legacy H-Sphere platform due to the end of PHP4 support by php.net and also security improvements in PHP5.

We’ll be conducting this upgrade on Monday 14th Jan 2008 at 23:00.

Please ensure any scripts you have are up to date with the latest versions to ensure compatibility with PHP5. After this date, we will not be able to offer any PHP4 access as H-Sphere only operates with a single version.

If you have any questions or concerns, please contact support.