News updates
|
August 15th, 2014
at 9:50 AM CDT |
| Finally back online after that Great Storm of
August 2006. The storm caused several changes that I'm still
investigating, including: server upgrades, server relocation,
firewall changes, virtualization, full redundancy, automated
off-site backups, capacity upgrades, bandwidth upgrades, power
upgrades, software upgrades, etc. Very bizarre, I'll try to post
an update with my findings before 2025.
|
|
August 20th, 2006 at 10:30 AM CST |
| A recent storm knocked utility power offline for
several hours this morning. The server was shut down after the
UPS runtime limit was reached. There was a total downtime of
approximately 3 hours.
|
|
August 12th, 2006 at 4:50 PM CST |
| MDaemon was updated from version 8.1.5 to 9.0.4.
This update includes performance enhancements across the board,
as well as updates to core spam filtering components.
|
|
March 6th, 2006 at 7:06 PM CST |
| All core services were moved to a new server and
a new IP. This new server is significantly faster and more
capable. The migration should be smooth.
Also, MDaemon was updated from 7.2.2 to 8.1.3. Most significantly, this update should allow for more advanced
|
|
January 1st, 2005 at 3:44PM CST |
| It has been a great deal of time since my last
news post. In that time, many things have happened. First, the
webserver issues were fixed with a change in software to Apache
2.0/win32. I also have upgraded the mailserver (from MDaemon
6.8.5 to MDaemon 7.2.2) software, and it now offers much more
robust webmail and vastly improved SPAM filtering through use of
SURBL, accompanied with an upgrade of the underlying
SpamAssassin software from 2.55 to 2.64. In addition, I now
require use of either POP-before-Send or SMTP-Auth before you
can send outgoing mail, and encrypted POP3, SMTP, and IMAP4
services are available, as well as SSL-secured webmail at
https://webmail.xfury.net:3002.
|
|
August 28th, 2004 at 11:42PM CDT |
| Once again, the provider has modified speed
profiles. Outgoing bandwidth was just increased by another
50%...a good 3 times what it was a year ago at this time, all
for free. Anyway, the problems with the webserver continue to
get worse, but the low traffic of this site makes it something
of a low priority. I'll install a different server sooner or
later. Spam control is getting better with every learn session,
and there should be a big learn next weekend.
|
|
November 10th, 2003 at 2:35 CST |
| The upstream provider has modified the speed
profiles of all customers. As a result, outgoing bandwidth has
been doubled. This should make the server considerably more
responsive, especially for numerous messages or attachments.
|
|
August 20th, 2003 at 4:41PM Central |
| When I attempted to upgrade the hardware and
software in the server as per scheduled outline, the boot drive
failed. The system has been reloaded fresh, and all webpages
were lost in the drive failure. All e-mail and e-mail accounts
were preserved, as they were on a different drive. All FTP
accounts were lost also, and these should slowly be coming back
online. If yours isn't up yet, e-mail me and we'll get it
figured out.
|
|
October 21st, 2002 at 5:51PM Central |
| Beginning sometime this morning and lasting until
10:00 A.M., server connectivity was intermittent. This was
caused by a rogue program consuming all CPU cycles and
preventing running daemons from picking up socket connections.
The issue has been addressed.
I'm planning to update the IRCd software sometime in the near future. This should result in faster connect times and fewer drops. Thank you for your patience.
|
|
October 14th, 2002 at 5:22PM Central |
| Some time ago, MDaemon 5.0.6 was updated to
6.0.3. Stay tuned for an update to 6.5.0. Recent audits of the
spam filter have continued to show incredible effectiveness
along with very low false-positive rates. All seems to be going
well.
I've decided to switch Xitami to Apache+PHP. I have no timeframe for this upgrade, just that it's planned. The IRC server is acting up and dropping connections for unknown reasons. I'm looking in to the issue.
|
|
May 9th, 2002 at 6:51PM Central |
| MDaemon 5.0.5 was updated to 5.0.6 today to close
4 vulnerabilities recently discovered. This upgrade should have
no effect on any account. If you have problems, please e-mail me
at the address in the previous news post.
I'm considering switching webservers from Xitami to Apache with PHP mod. This upgrade would allow users to utilize PHP in any websites hosted here. Thanks!
|
|
May 7th, 2002 at 7:06PM Central |
| Due to several various issues with the mail
server software, I downgraded from MDaemon 6.0.0o Beta to
MDaemon 5.0.5 Release. This should fix problems with the webmail
interface not responding.
When MDaemon 6.0.0 goes gold (final stable release), I'll look into upgrading once more. If you have problems, please e-mail me at Admin (at) xfury.net Thank you for your patience.
|
|
February 25th, 2002 at 4:09PM Central |
| Previous mailserver upgrade took place during the
scheduled dates without incident.
On Sunday, February 24th, 2002, corruption in the array's allocation table took out the drive. The automated backup, dated 2/24/02 at 3:00AM, was restored. Therefore, minimal loss should have occurred. This problem has been reoccurring infrequently from the start of the server, and measures have been taken to insure that it never happens again. After restoration, mailserver software was upgraded from 6.0.0n to 6.0.0o. This is a minor upgrade. Thanks for your understanding.
|
|
February 2nd, 2002 at 12:27PM Central |
| The mailserver software is scheduled to be
upgraded next weekend (Feb 9-10). There may be periods of
inaccessibility within that time. I will update this page when
the transition is complete.
The major improvements are: -Enhanced interface for the Webmail users (I will be able to customize the interface) -Integrated antivirus software to keep mail free from infection. Stay tuned.
|
|
January 27th, 2002 at 2:09PM Central |
|
An upstream (redback) router owned by Southwestern Bell was down for a couple of hours this morning. Remote mail should be back queued and delivered shortly. Furthermore, due to changed routing, speed has been less than optimal lately. Hopefully with this redback router fix, speeds will improve.
|
|
January 14th, 2002 at 2:03AM Central |
|
Automated (scheduled) backup software (Freeware) installed. We haven't had disk issues ever since I broke the RAID-5 array into a RAID-0 array, but it never hurts to keep a comprehensive backup. I'd sure like to know why there were such issues with RAID-5. It was a Mylex DAC960 (read: solid hardware) with 10,000rpm IBM SCSI SCA disks. I don't quite get it. The 36GB array is now backed up to the Maxtor DiamondMax 60 drive to insure data integrity.
|
|
January 13th, 2002 at 2:20AM Central |
|
Downtime on Sprint routers in Kansas City took down access for approximately 30-45 minutes Friday evening. There were DOS (Denial Of Service) attacks several months ago against the server which affected availability. However, I'm just now getting around to updating news. :)
|
|
October 9th, 2001 at 10:27AM Central |
|
Firewall rule corruption took down all access to the machine. This has been repaired, and I'm looking into a new firewall solution.
|
|
September 26th, 2001 at 7:57PM Central |
|
Xitami 2.5b4 was upgraded to Xitami 2.5b5, fixing the directory list issues. Any new problems should be reported to the webmaster. Thanks.
|
|
September 24th, 2001 at 6:57PM Central |
|
What an utter nightmare. First, all damaged equipment has been replaced. This includes the primary switch, and (3) Intel 82559-based 100-Base-TX Ethernet adapters. Also damaged was port #1 on the 9-port 10-Base-T hub. However, since the hub is only running at 33% capacity anyway, this is not a big issue. (Switch1, on the other hand, operates at 100% capacity, and Switch2 at 34% capacity, soon to be increased to 62% capacity) Second, the web daemon software was successfully changed to Xitami. And finally, the HD volume that hosts the mail server, the Port remapper, the Triviabot, and ircd, as well as local file storage, crashed. This also has been repaired, and efforts are underway to get a new drive to run daily automated backups to. Anyway, everything should be fully functional now. If you have any problems, please contact me at Admin (at) xfury.net
|
|
September 18th, 2001 at 6:28PM Central |
|
Switch has not been replaced yet. However, there will be some WEB-SERVER downtime on Saturday (Sept 22nd) during the late AM or early PM hours as I change web daemon software. Microsoft IIS 4.0 is riddled with security holes, and I just don't have the time to mess with patches every 2 days. The new software affords some very interesting options for those wishing to host here - stay tuned to find out. Note: All content and functionality should remain.
|
|
September 18th, 2001 at 1:23PM Central |
|
The cards have been reported by the retailer as approved for shipping today. They should be in tomorrow. Installation of these cards should not affect connectivity for anyone until Saturday, when Stargate will be taken offline for it's primary card change out. (Stargate is currently connected to the network solely by its secondary card in a performance and connectivity-degraded mode)
|
|
September 18th, 2001 at 12:29PM Central |
|
The new replacement switch has arrived and will be installed during the evening hours. The Ethernet cards had a problem in the order and -hopefully- will be here by tomorrow. There may be some unexpected, abrupt downtime tonight as the new switch is brought online, but it should not last more than a few minutes.
|
|
September 17th, 2001 |
|
Early this morning (Sept 17th), lightning struck the network. While most machines are protected by some varying model of APC UPS, -not- all are. It is my suspicion that lightning entered the network via one of these unprotected machines and dispersed through the network, causing equipment damage. Known dead equipment: (1) Netgear FS-108 switch (Primary switch) Suspected dead/damaged: (1) Intel EtherExpress PRO/100+ Management NIC Replacement equipment has been ordered and shipped UPS Next Day Air. The network is currently running in a performance-degraded state with many nodes completely unable to connect. There may be unexpected service interruptions until everything can be replaced.
|