Forum Archive Index - May 2002

[Adam Rands <adam@iguana2.com>]

As most of you may know Iguana2 hosts and manages the Access Brokerage website.

I have been asked by Access Brokerage to inform you about the outages today 
and other recent events causing problems with reliability.

*** warning - slightly technical details follows ***

Today at 11:50NZT the Iguana2 database machine suffered a memory failure.

This failure should of been automatically handled by the machine as it has 
*chip kill memory* (essentially RAID for memory). It should of unallocated 
the faulty memory SIMM and continued with its work.

Unfortunately this did not happen and the machine rebooted (and corrupted 
the system disk in the process). We will be asking IBM why this happened 
when they come out to replace the faulty memory (which will be done in the 
small hours sometime this week)

The database machine itself is a IBM NetFinity 5500 with full RAID, 2 power 
supplies, intelligent (sic) RAM etc. It is specifically built to handle 
faults on the fly and *not* behave destructively. We could not have 
foreseen or perverted today's problem.

As soon as we had diagnosed the problem we had to rebuild the system disk 
so we could access the Database (on RAID). This was what caused the lengthy 
downtime. (approx 2.5 hours). Given the work involved its a testament that 
our tech team managed to complete  this work as quickly as we did.

Many of you will also be aware that there were delays with the NZSE market 
information on the site yesterday. These problems were the result of 
Iguana2 implementing the NZSE new broadcast system over the weekend. I 
don't want to comment in detail on this issue but needless to say the 
problems was out of our hands.

The other issue effecting reliability over the month has been 2 unplanned 
10minute Internet outages.

We connect to the Internet via NTT (the worlds largest Telco). The two 
recent outages have been on the Australian backbone and affected 80% of 
*all incoming traffic* to Australia. Once again we are powerless to prevent 
these sorts of outages. We simply have to log a call with the NTT (along 
with the 1000's of other companies within Australia) and wait until their 
engineers correct the problem.

Why am I telling you all this?

I just want to reassure you that Iguana2 and Access Brokerage are taking 
all steps necessary to ensure the on going reliability of the site. Today's 
downtime was the longest period of downtime in Iguana2 six year history and 
not something we care to repeat. We pride ourselves on our uptime.

I apologise for the problems today. Rest assured that we do attempt to 
prevent issues like this occurring and are committed to improving the 
reliability of the site (where we can) in the future.

Also......

In the very near future Access Brokerage is launching a completely new 
website. If you think StockNessMonster is a good site then I am 100% 
positive that you are going to be well impressed when Access launch their 
new website (built by Iguana2)

If you want to discuss this matter further please email me off the list


Adam Rands
COO, Iguana2
adam@iguana2.com


----------------------------------------------------------------------------
To remove yourself from this list, please use the form at
http://www.sharechat.co.nz/chat/forum/