Our technical support team will be working normal coverage for the holidays, both on Christmas and January 1, 2012. Billing and accounting will return on December 26th and January 2, respectively, from the holidays.
From all of us here at Hosting Matters, to all of you out there, we wish you a Merry Christmas and a happy and most prosperous new year. We look forward to having you on board in 2012 and beyond.
On Saturday, December 3, beginning at 8 PM (Eastern) we will be migrating all monet-based accounts to the new replacement monet. If you are managing your own DNS at a third party location, please contact us so we can provide you with the new records to enter for your site(s). Please do not make any changes to third party DNS records until the migrations are complete, as this will render sites and/or email unavailable. If you are utilizing the serverwide certificate on monet, no changes will be required as the server name will remain the same.
If you have any questions related to this, please contact support via the usual means.
On Saturday, December 3, beginning at 8 PM (Eastern) we will be updating the client system. This update will bring with it a new look and more intuitive controls on the client side, along with certain updates on the back end for us. All current profiles will be unaffected by the update and will remain valid once the update has completed. The client system will be unavailable for less than five minutes for direct login during the update routine. Support requests sent via email will still automatically generate helpdesk tickets during this period.
If you have any questions related to this update, please contact support via the usual means.
On Friday, December 2, beginning at 8 PM, accounts on wyeth, mondriaan, and magritte will be migrated to new servers. Sites on mondriaan and wyeth will be relocated to the new mondriaan, and sites on magritte will be relocated to the new magritte. Those of you managing your own DNS elsewhere will need to update IP settings at those locations; if you are managing your own DNS, please contact us so the new settings can be provided to you. Please do not make any updates until the migrations are being done (or have been completed). Doing so will render functionality tied to the site(s) inoperative. If you have a site/sites on wyeth, and use the serverwide certificate, references to wyeth will need to be changed to mondriaan once the copy process is complete.
The older servers and the new servers are running the same mySQL and PHP versions, so there are unlikely to be any issues with applications as sites are copied over. However, if you do encounter any issues, please let us know immediately so we can help get them worked out.
If you have any questions about this, please contact support via the usual means.
One of the drives in the RAID on mucha requires replacement. That will be done today. Total impact time will be a couple of minutes to power down the server and swap the drive out. If you have any questions, please contact support via the usual means.
Thursday, November 24, is Thanksgiving in the US. We will have coverage throughout the day, and will be handling technical support issues and monitoring as usual. Billing staff will return on Friday, November 25.
We hope everyone has a happy – and safe – Thanksgiving, for those observing it. For everyone else, have a great day heading into the weekend.
This afternoon we suddenly began experiencing issues with the nike server. Based on the log outputs and the output of the filesystem check that was required when the server was rebooted, the planned retirement date for nike, scheduled for the end of the year, may be stepped up to an immediate migration of accounts and retirement of the hardware. At the moment, we are attempting to get the filesystem to a consistent point to allow for graceful migrations to other servers of the handful of accounts that remain on nike. If that ends up being unworkable, we’ll pull the backups and restore from the last backup rotation to other servers instead.
If you have any questions related to the issues with nike, please contact us via the usual means.
Retirement time has come for the next two servers remaining from the older server group. Both rockwell and gerard are being replaced by newer, faster versions of themselves, and will retain the same server names. Migrations will begin on Friday, October 28, at 8 PM Eastern.
For residents on both servers, if you are managing your own DNS at a third party location, please contact us (with your domain name) and we’ll provide you with the settings to input at that location to point the site(s) to the new server. As the server names will be the same, no changes are necessary if you are utilizing the serverwide certificates.
We will copy all content for each site between servers, including web content, databases, and mail.
If you have any questions related to the retirement parties for these two servers, please let us know.
We are currently tracking an issue with the mailscanner process not scoring any mail passing through the queues. Based on a review of the logs, the antivirus portion of the scanning is working without issue; however, the logging also indicates an error within mailscanner that it believes the spam scoring module is not present, although it is and can be tested manually to confirm it is available. Thus far we have confirmed this activity on calypso, mucha, longhi, and hermes. We are certain others will follow since this tends to be a cascading type of error as the configuration is identical across servers. The result of this issue is all mail, beyond mail that is caught by the antivirus as infected, is scoring as zero and passing directly through to all mailboxes.
Work on this issue will continue until it is resolved. At this time, there is no ETA. Updates will be posted to the forums as well as to the remote status page at http://hmstatus.com
UPDATE: We have tracked down the underlying issue and a fix was tested on calypso and longhi. The repair was successful, and it is now being rolled out to all servers, regardless of whether or not they are currently showing this particular problem.
The weekend of October 8-9, we will be rolling out a maintenance update to the client area application. All functions in the system will remain active during this process and no client impact is expected. This maintenance release is primarily for certain backend functions and some APIs.
If you have any questions related to this maintenance, please let us know.
