FQuest Alert: SIX downtime

[Terra]

At 2:54pm EDT, SIX locked up *again*...

The server was revived and running at 3:32pm EDT...

I have been on the phone **again** today with my suppliers, and should be receiving the new server very soon...

Due to the specialty equipment that we use, product availability is at times very difficult to obtain... The memory that I received today was the wrong type, so I am scouring the nation to find another supplier that has the particular high-speed Toshiba memory in stock and have it red-labeled in... I have also been in touch with Tyan and they feel that it is a combination of both motherboard (FAN problems) and memory problems...

Our goal is to have a unused standby server always available, to handle situations such as this...

SIX has proven to be unreliable and problematic, yet it is using almost the same hardware as TAZ is... I have also been in touch with the Linux core developer team and has been determined that it's definitely hardware and not the Linux 2.2.x kernel...

The downtimes with SIX has caused me to take a serious look at our redundancy problems, and up until now having a non-production spare server has been cost prohibitive...

We are working diligently on this problem and hope to have it resolved very soon...

Our sincerest apologies for the problems that SIX has caused for everyone... I just hope someone will pinch me soon to wake me up out of this nightmare...

--
Andrew Gillespie
Systems Administrator
FutureQuest.net

[Charles Capps]

Always remember things COULD BE WORSE...

Question: "FAN" - is that an acronym, or is it just a fan? Honest question...

------------------
"Okay, so I'm not "SANE" so to speak, but uh... I'm the lovable kind of psycho"
http://solareclipse.net/

[Terra]

I have ***finally*** tracked down the specific Toshiba memory from a Dallas supplier...

Getting memory is very volatile, just like the stock market... The memory I got today, was the last 8 available, and they only received them in late last night... Oh the joys of procurring high-tech server equipment...

For reference, we use Toshiba TC59S640BFT-80 pc_100 SDRAM ECC chips... TAZ has run flawlessly with this memory, and hope that SIX will follow suite when I get rid of the Micron memory chips now installed...

I am also obtaining 4 P-II 450 CPU's today as well, in preparation of taking the SIX server SMP capable (dual-cpu)...

All future servers will be SMP capable... This should hold us till the AMD K-7's are released in .18 micron form... http://www.aota.net/ubb/smile.gif

We hope that this will help our redundancy somewhat until I can begin building the clusters...

Charles:
yes, FAN was in reference to the motherboard CPU FAN connections... When the fans shutoff the last time, I hardwired the fans directly into an unused power supply floppy drive connector, instead of relying on the motherboard connection... I thought that it was due to a problem with the APM (power saving) features --which were all turned off in the BIOS-- and somehow this was shutting the fans off as well...

--
Terra
--Which server do you want to kick today?--
FutureQuest.net