Eskimo.Com Shell Up

The old eskimo.com shell server has been restored to service.  It wasn’t completely crashed.  SunOS 4.x has a strange bug in it where it can get in a state where the kernel is still running, the machine will still respond to pings, but any user space programs don’t get CPU time.  This seems to depend upon the CPU module, the previous module never did this but it cooked, this one does, and I’ve no longer got any spare parts to replace it with.

Classifieds…

     I’m considering adding a classified section to Eskimo’s website.  Maybe limited to tech only (jobs, equipment, tools, literature, etc), or maybe include music, art, and a few other things but generally limit scope in order to prevent it from becoming a scam, drug, and prostitute classified service like some others have become.

     The reason I’m considering this is that I frequently have both customers looking for employment and people approaching asking for some of these things so this would be a way to get them together.

     I’d like your input please.

Help Me Grow Eskimo

Please help me grow Eskimo!

     There are huge economies of scale, more customers will mean a lower cost per customer.  The fixed costs of operating Eskimo are significantly larger than the incremental costs.

     Presently, there are hosting providers out there that are considerably cheaper, but they achieve this cheapness by putting 10,000 domains on one box that is running versions of Linux two major point releases behind.  They tend to run all websites under the same user ID because it’s less complex and provides better performance on their overloaded and out of date platforms. I won’t mention any names but their ads tend to either feature cows or bikini clad young women in sports cars.

     I’d like to be able to provide our services at those costs but with much better performance and security that we presently provide. Actually I’d like to improve both even more.

     There are things we can do on a larger scale that just aren’t doable on this scale, and the things that are doable on this scale can be more optimized.

     For example, I’ve got the web server now to where it serves most pages in under 200ms, this is important from a Google ranking perspective and helps make your pages more visible on the web.  But getting it down even further would be possible with higher end hardware.  For example, the motherboards I’m using now have a maximum memory capacity of 32GB.  The more memory available, the more things can be cached and the less hard drives need to be depended upon, this speeds things up.  I’d like to go to a higher end motherboard that can hold 256GB or even 512GB of memory, but too costly on my current budget.  I’ve avoided solid state drives in the past because once all the banks are full they tend to become slower than rotary disk drives and have data errors, however, there are some server grade drives now that have been tested at over three petabytes read/write and remained fast and reliable but the costs tend to be $800+ on these.

     I’d like to further develop our site for mobile use and integration but that’s a time consuming prospect that really requires additional bodies.

     Most modern businesses these days seem to devise some product and then figure out a way to market it and convince people, who never previously had a need for it, that they just can’t live without it.

     I’ve always taken the opposite approach, listened to my customers and their needs, wants, and desires, and tried to come up with products that serviced those needs and wants.

     In order for this to work I need your input, what it is that you want or need.  The amount of feedback I’ve gotten hasn’t been strong recently.  People tell me they’re happy with the service, in deed few are leaving except by death, but at the same time I’m not being overrun by hoards of new customers.  So please tell me what you want and what you think Eskimo needs in order to not merely survive but to thrive and grow.  Please tell your friends about us.

Ubuntu Updates are BROKEN!

Ubuntu’s latest updates deleted freenx, x2go, my pre-emptive kernel, and nividia drivers leaving my workstation totally useless, text only at present.

I’m trying to get it operational again, a necessary prerequisite to fixing the other things that are broken like vnc and the x2go webpages.  Your patience is appreciated.

VNC / Shellx

VNC seems to have died on shellx, I am only getting a blank black screen and cursor when I try to connect to it.  I highly recommend x2go as an alternative, it’s faster, smoother, and supports sound.

I’ll work on troubleshooting VNC but it’s been so many years since I set it up, and it’s such as slow kludge of a program that I haven’t used it in as many years, so it’s going to take some research.

Fedora Up

     Fedora is back up.  There was nothing in the logs that gave any hint as to why it rebooted or why it started the emergency shell.

Starting Emergency Shell…
[ OK ] Started Emergency Shell.
[ OK ] Reached target Emergency Mode.

Fedora Down

     The Fedora 21 shell server is presently down.  Sometime during the night it rebooted into “emergency mode”.

     I am in the process of troubleshooting and working on returning it to service.

Again Mammoth Maintenance Not Done

     Again, Mammoth DSL maintenance that was supposed to have completed by 8AM mountain time (7AM Pacific) is still in progress.  It is taking longer for them to swing the circuits than anticipated.

     This effects most Seattle DSL customers on CenturyLink DSL circuits.  They are working to restore out of service circuits as fast as possible.


 

Date: 3/11/2015
Start time: 5:00 am MST
End time: 8:00 am MST
Affected: 74/OBGJ/090579//ACSO

Detail:

Extended: Maintenance is being extended in order to complete the OC3 Groom.
Most customers will see a short drop 30 seconds to 1 minute and restore to
service. As usual some customers may need to reboot to restore service.

Maintenance is being performed in order to groom the OC3 circuits to the new
OC12 circuit. All ATM customers on: 74/OBGJ/090579//ACSO will be affected.
Customers will go down and come up as their circuits are groomed to the new
service, some customers may need to reboot to restore services. Estimated
downtime is roughly 2 hours but maybe as short as 5 minutes for some
customers.