Bridge Problems

here is Jennifer's message ....
Sheri
---------------------- Forwarded by Sheri Thomas/HOU/ECT on 09/12/2000 10:14
AM ---------------------------

From: Jennifer deBoisblanc Denny
08/15/2000 02:43 PM

To: Beth Perlman/HOU/ECT@ECT
cc: (bcc: Sheri Thomas/HOU/ECT)
Subject: Bridge Problems

Beth,
I know you are at the VP level but I thought I would forward you the memo
below.
I wrote it to explain to the EnronOnline Call Center the various issues we
have been having with the "bridge" since they get a vast
number of internal calls when we have a problem.

Please let me know if you have any questions or if I should be funneling
these concerns to a particular person.
Thanks.
Jennifer
3-5824

---------------------- Forwarded by Jennifer deBoisblanc Denny/HOU/ECT on
08/15/2000 02:31 PM ---------------------------

From: Jennifer deBoisblanc Denny
08/10/2000 05:22 PM

To: EOL Call Center
cc: Sheri Thomas/HOU/ECT@ECT, Frank L Davis/HOU/ECT@ECT, Torrey
Moorer/HOU/ECT@ECT, Dale Neuner/HOU/ECT@ECT, Tara Sweitzer/HOU/ECT@ECT
Subject: Bridge Problems

Over the last 8-10 business days, we have had a some issues on the bridges. I
want to educate you a little bit on what can cause
"bridge problems".

First, what I want to explain is their is ONE GUI bridge application for
users (backoffice users) to monitor the bridge but there are MULTIPLE
systems that EnronOnline transactions can be pointed to downstream. This is
important to understand because if you do not know this ,
it may seem like we are having the same problem over & over; ie "the bridge
is having problems". But in actuality, it is DIFFERENT
problems. Still not good - but at least it cannot be isolated to ONE
problem.

Also, due to multiple systems (Sitara, Enpower, GasDesk, Tagg/ERMS,
Powerdesk) are listening for messages from one EnronOnline,
a "bridge" problem could be one of many issues. Some of which are:

The server for EnronOnline messenger goes down (doesn't send message to
bridge to process)
The Enron network goes down (cannot send message thru bridge)
The NFS server (Enron main server) goes down
Sitara server goes down
Enpower server goes down
Tagg/ERMS server goes down
Deal worker in ANY system gets "hung up"
Huge volume slows down processing time for one or more downstream systems
Scheduled outage for maintenance on server on weekend was not re-started on
Sunday

and the list goes on.....................................

FYI In the past week & 1/2, the following actions happened, which caused
"bridge" problems.

Scheduled outage for maintenance on TAGG/erms server on weekend was not
re-started on Sunday (not smart)
The NFS server (Enron main server) goes down (affected Sitara & Tagg/ERMS)
Deal worker in Sitara system got "hung up" and servers had to be bounced
Member of EnronOnline Dev Team changed data for specific tokenised periods
during trading hours instead of end of day
(This failed deals in bridge for Sitara & Tagg/ERMS)

One thing to remember is you are getting calls from traders & back office.
Traders are calling because they cannot see deals in their
position manager application of their system. If they are gas traders, if
could be Sitara or Tagg system. If they are a power trader
calling, they cannot see it in Enpower. ecetera, ecetera

If a back office person calls, they are watching the monitor GUI and they can
see deals queing/backing up and know there is a problem.

We are working to resolve some of these issues, particularily,
multi-processing tools, as we know the # of transactions will only increase.
Please feel free to email me if you have any questions. I hope this was
helpful.

See below for explanation of problem on Wednesday.

---------------------- Forwarded by Jennifer deBoisblanc Denny/HOU/ECT on
08/09/2000 03:50 PM ---------------------------

KARIE HASTINGS
08/09/2000 02:06 PM
To: Jennifer deBoisblanc Denny/HOU/ECT@ECT
cc: Mark Kinkead/HOU/ECT@ECT, Scott Mills/HOU/ECT@ECT, Thomas
Engel/HOU/ECT@ECT
Subject: Sitara Production Problem

I am forwarding a message that explains the Sitara Production problems from
this morning. Please let me know if you have any additional questions.

Karie (36759)

At 9:50am this morning a DealWorker process in the Sitara system became
unresponsive.
At 10:00 am it was determined that the EOLBridge in sitara was not
processing new deal requests
and diagnostics lead to an indication that the dealworker was not functioning
properly.
At 10:14, am the Dealworker was forced down and a new one restarted.
At 10:21 am , the EOLBridge began processing deals.

The cause of the is difficult to determine, but indicators point to an error
in the CORBA product we are using,
most likely due to scalability issues of that product and of Sitara. This is
an issue that is actively being addressed.

--Mark Kinkead
3-1961