Aphrodite is currently experiencing a large WordPress attack.
We have attempted to mitigate the problem using mod_security, but the attack is so large that we’ve taken the step to disable WordPress logins for the time being. If you are trying to login to WordPress you will see an error that says “Not acceptable”.
We will keep you updated with any developments.
19:32hrs – ceres.uksrv.co.uk reported a kernel problem. We investigated and tried to correct the issue, but it became unrecoverable. The server was rebooted at 19:49 and we are bringing services back on incrementally to stabilise the server.
19:59hrs – ceres.uksrv.co.uk returned to operational status.
13:45 A large volume of malformed requests started to land on our network entry points. At the moment our systems have been overwhelmed by what appears to be a focussed attack on our network. We are working hard to mitigate the issue.
17:00 All systems have now been operation for quite some time, please contact support if you are having any issues. We are continuing to go through our data and take steps to mitigating any further attacks.
Around 23:00 on the Sunday the 17th of March we’ll be performing some essential network maintenance that governs access to the internet for dedicated servers.
We have carefully planned the work and already successfully performed it on the all other server clasess, e.g shared, premium and reseller and VPS so are confident that it will be non-disruptive.
We anticipate no noticeable downtime (a few dropped packets perhaps but nothing service affecting)
If you find that your server has connectivity problems for more than 5 minutes please call us on 03333 44 1337 choose option 3 and use your out of hours PIN that has been recently emailed to you.
We’ll update this blog when starting, during and upon successful completion.
23:30 We’re now starting this work
01:15 A number of servers are actively using IP addresses outside of their allocated ranges so we need to postpone this body of work until we can migrate these IPs. Thank you for your patience.
[07:00] This morning at 7:00hrs eris underwent an unavoidable reboot due scheduled kernel updates. Unfortunately, upon restarting, the hardware reported an on-board memory failure. Two further cold boots were attempted, but unfortunately eris would not start. The system did not report any hardware/firmware errors, so this failure was not anticipated. We are currently investigating the issue at the datacentre and hope to determine which memory module is causing the issue soon.
[07:58] The bad memory module has finally been identified and removed from the system. We will carry out some analysis on the system today and refit new memory this evening or later, which will require a few minutes while the system restarts again. This article will be updated when we schedule that event (we do not anticipate any further issues).
On the 1st of March between 23:00 and 01:00 all shared servers will be moved to their final homes as part of our datacentre consolidation.
We anticipate up to 10 minutes of downtime per server but hope it’ll be closer to 5.
Thank you for your patience.
All email will be held by StriKe and forwarded when the server is back up.
00:17 This work has been completed successfully. Thanks for your patience.
As previously communicated (In October, December and this week) the hardware migration of Artemis is starting at 5pm on the 10th of January.
This is scheduled to take up to 10 hours. All emails sent to you will be held by StriKe and forwarded after the move – no email will be lost.
Thank you for your patience during these essential works.
We will keep this posted updated, please check back.
[17:31] – Old Artemis has been disabled
[20-:58] – Work continues on the migration
[21:58] – Almost all of the databases and bare accounts have been moved over, we will be starting moving the /home partition in the next 30 minutes.
[22:49] – /home directories are now syncing.
[00:04] – /home still syncing
[01:34] – /home is synced, we are now configuring services.
[02:15] – Artemis MK2 is online and serving!
If you are having any problems with your website then it’s almost certainly due to the upgrade from php5.2 to php 5.3.
Please investigate the error_log file inside your public_html directory. If you are till having problems then please contact support.
Update: January 11th
We are aware of an issue to clients outgoing mail servers. We are currently in the process of resolving this.
If you are experiencing any problems of this nature then please submit a ticket.
Update: January 11th 12.20pm – All email issues should now be resolved.
[This work has now been completed, thank you for your patience.]
Between 22:00 and 23:59 on the 8th of January we will be upgrading the kernels on all shared hosting machines.
We expect downtime of no more than 15 minutes per machine.
Please check back here for status updates as the work progresses.
This work is essential to ensure the continued security of the fleet, thank you for your patience.
Aphrodite – Completed – Started 22:33 Finished 22:43
Apollo - Completed – Started 22:33 Finished 22:43
Ares – Completed – Started 22:41 Finished 22:53
Demeter – Completed – Started 22:51 Finished 23:02
Dionysus - Completed – Started 22:51 Finished 23:01
Hera – Completed – Started 23:00 Finished 23:11
Hermes – Completed – Started 23:00 Finished 23:12
Hestia – Completed – Started 23:14 Finished 23:27
You can find your server by going to <YOURDOMAIN>/cpanel – the server name is before krystal.co.uk
What?
Servers Sahara, Pluto & Ceres are being moved between chassis in the datacentre as part of consolidation works.
When?
The servers will be powered down and safely moved after midnight on Wednesday night. (So very early Thursday morning)
How long?
We anticipate 10 minutes of downtime per blade as it’s shut down, moved and powered on.
What does this mean for me?
Whilst the server is down your website and cPanel won’t load.
E-mail will be held by StriKe, our spam cluster and re-delivered when the system comes back up.
The machine will be shut down cleanly and no data will be lost.
20th September 03:03 – this work has all been completed.
What?
Over the last few weeks Apollo has had a number of a small outages lasting 5 to 10 minutes. After extensive discussions with Dell, the hardware provider, and Cloud Linux, the operating system vendor, we are at the point where we need to take the server offline to upgrade the firmware to eliminate it as a potential cause.
When?
We are planning to start work at 00:00 GMT on the 14/09/2012
How long?
We expect this work to take up to 2 hours
What does this mean for me?
Whilst the server is down your website and cPanel won’t load.
E-mail will be held by StriKe, our spam cluster and redelivered when the system comes back up.
The machine will be shut down cleanly and no data will be lost.
We apologise for the inconvenience this causes. This isn’t a step we are taking lightly.
[00:17 Friday 14th] - Preparing the server for offline mode now, expect everything to go offline shortly.
[00:28] Firmware update has begun
[01:18] Firmware upgrade complete, server is coming back online.
