More FurryMuck Status Reports
FurryMuck is indeed up, but their temporary lines are being overwhelmed by addicted muckers. No folks, FurryMuck cannot be run on much less than a dedicated T1 or equivalent. :-) Read below for a couple of updates from the site admin, S'A'Alis.
As we've posted here before, here is the situation:
On Thursday, May 10 at approximately 2:38pm CDT, the regional internet
provider for FurryMuck (SM) declared insolvency. While we had redundant
lines,
this cut *both* services as this was the largest DSL provider in Northeast
Wisconsin and it took down all others related to it. Further, our routes
were lost, as our IPs were also removed due to this situation.
Our team worked late into that evening getting the server put up on a
temporary network address. We are still having problems with DNS and
websites due to technical issues stemming from the methods used to
get FurryMuck (SM), itself, back on line. So we have opted to, for the time
being, get FurryMuck (SM) running again on an IP address and worry about the
websites and DNS later.
Yesterday, Friday, we declared the FurryMuck (SM) server available at noon
CDT. The current address is:
12.37.14.197 8888
This morning, we are working to restore DNS, mail and web services as
well.
Three weeks ago, we started work to have installed two new lines from
different (completely separate) providers using completely separate
technologies, to insure this does not happen again.
Unfortunately, due to "lead times", the soonest the installation of the
first of these could be completed would be Friday, May 18, 2001. We have
contacted that provider and are trying to expidite that installation. A
site survey is occurring Monday morning and we *hope* to have installation
by Wednesday, but they can't guarantee us of that.
Because, however, of the amount of work involved in changing to a completely
new layout and scheme in order to use our temporary solution for FurryMuck
(SM),
email, web and DNS, and then another completely new scheme on the new
installations (due to the fact that they are disparite technologies), we
have
decided to prioritize our services. That is, we will make sure the Muck
itself is running. Then we will work on DNS, then email, and finally web.
If some of these don't get done by the time installation of the new lines
at our home site begins, we will not continue to work on those services but
will wait for installation to be complete so that we can focus completely
on our new lines.
The upshot is this -- we are focusing ALL our efforts to keep the Muck
itself
up and running and as available as possible.
We have had to do this temporary measure with alot of haste and it is not
the best possible solution, but it DOES insure that FurryMuck is at least
available, if a bit slow.
We would like to thank the crew at FugaTek for their assistance and donation
of their site, services and expertise to push this through post haste.
We will continue to post messages here and, via the wizzes, to other boards
and Mucks until this situation is completely resolved.
The second of our new lines is due to be installed in early June.
We apologise for these problems and hope you understand that the
circumstances
were completely beyond our control and that we *are* trying to do everything
possible to get all services back as quickly as possible and with a minimum
of "pain and suffering".
-S'A'Alis
And from late Saturday morning:
The temporary lines are so overwhelmed, currently, with people trying to
get into the server, that NO ONE is able to function on FurryMuck (SM) at
this time.
IF YOU CAN REDUCE YOUR TIME ONLINE IN ORDER TO ALLOW OTHERS IN TO CHECK
THEIR PAGEMAIL, ETC., PLEASE DO!
We are working to get all functionality restored, but the temporary lines
can't handle this load. Please be considerate of others and help us out
here.
We're working as fast as we can and would appreciate your cooperation
and understanding.
-S'A'Alis
About the author
Aureth — read stories — contact (login required)an agronomist and Cornwuff from Northern Illinois, interested in sf, homebrewing, photography and running
Comments
On a totally off-topic remark, it's interesting to see the use of "(SM)" (service mark) popping up in everything related for FurryMuck (SM). It's in these news updates, it's on the logon screen, it's everywhere. I guess since the wizcorp had to find some legal recourse with which to wrest away control of the furrymuck.* DNS domains and shut down the Ebay auction, they've decided they better keep using it and make it really obvious. Intriguing, and actually pretty amusing.
Unfortunately, that's how the law is, and how the ownership of the domains is resolved. They do have a few lawyers behind them now, which I think is a good thing.
The temporary lines are so overwhelmed, currently, with people trying to get into the server, that NO ONE is able to function on FurryMuck (SM) at this time.
I find this odd because FurryMUCK seems to have zero trouble getting as far as quickly sending the initial (pre-connect command) page. In the 1/3 of the time things go beyond that, everything seems nice and zippy until *thud* no more connection, not "slow".
It looks identical to a case I saw where someone made the mistake of turning on Explicit Congestion Notification in the 2.4.x Linux kernel when they recompiled the kernel, without bothering to check the warning in the config process, where it notes this does not work with many routers and firewalls yet. Though I've seen the same with an *extremely* overwhelmed and underpowered router.
I'm assuming of course that kernel limits on the number of connections, et cetra aren't in effect or have been turned off in /proc and PAM's config files.
I noticed that 3/5ths of those connected right now (Sun, May 13, 5:30pm) have idle times equal to their connect times. They must have tried to get in and been disconnected almost immediately. The admins might want to reduce the idle timeout to more quickly collect these garbage connections.
Many of us who have excess hardware located at internet
sites offering high ammounts of connectivity are surprised that
S'A'Alis won't accept offers for temporary hosting at key
internet POP's. I know myself, and several others can offer
significant server resources at places like MAE West, Digex,
BBN, XO, Excite, and many others! Why FurryMUCK has been
relegated to slow DSL service for so long is totally beyond
my comprehension. Personally, I have offered hardware, and service
many times in the past, and have had the offers rejected. Even
offers to donate a dedicated machine, and bandwidth.
SMASH ([email protected])
I suspect part of that is to prevent the existence of FurryMUCK from being beholden to any one person or group. I can see where they'd feel it puts them in a compromising situation. For example, if somebody "donates" hardware, what if the game then comes to depend on it and then the person wants the hardware back, or decides he wants some compensation, or tries to use it to leverage a "favor" out of the wizcorp?
Or maybe they're just being obstinant.
-Feren
"We use them for divine retribution."
Isn't FM (SM) *already* beholden to an individual who is keeping it on a DSL connection and being obstinent in the face of offers for temp hosting on fast sites while other arrangements could be made (i.e., S'A'Alis)? I don't know the situation there, if there's only one person calling the shots or a committee, it's *still* beholden to a specific group or individual that favored basically taking it offline for everyone except those of us who looked around and found the IP, in order to keep it on their hardware and connection.
However, I can also see reasons to reject offers--
it's their "baby", I know I don't like to just hand over any control whatsoever to just anyone who comes along, even if it would be somewhat of a help in the short term on my projects. There would always be the possibility (however remote)of abuse on the part of the "host" in regards to the privacy of the players, especially if the MUCK's data files are archived in plain text. There are also liability and other concerns that come with putting such a service on a commercial enterprise's systems and connection, especially if all the powers-that-be who are overseeing such things aren't fully aware of what's going on or granting appropriate permission.
If this had dragged on for weeks with no end in sight (sort of like the Furnation outtage), then maybe I'd start thinking it was a bit selfish to refuse offers of help that would benefit the huge userbase, but fortunately that wasn't the case and everything seems to be getting back on track.
--Pyrrhus
Post new comment