Hi Tim,
It might be more advantageous to update your Change Control
Procedures documentation to manually perform #3 before they
begin and #5 when they complete. These can be integrated into
a Standard Operating Procedure for all changes, similar
to calling the help desk before commencing changes, and
performing pre-change checklist activities.
I'm a fan of automation, but consider whether you are
building additional automated procedures that may
not apply in every situation- Change Window cancelled,
Change Window rescheduled, or that fail due to unexpected
conditions- database crash, network unavailable, etc.
I've found over the years that standardizing procedures that
people perform is just as important as finding ways
to automate things.
Just a thought.
That being said, I think the steps you've outlined below
mostly cover what you are attempting to do, but I'm
thinking an additional step has to happen at 3.5:
- Update the change control table with each completed item
i.e. set Compete=1 on each item.
Cheers,
Jim B.
________________________________
From: Tim Peiffer [mailto:peif...@umn.edu]
Sent: Tue 6/8/2010 11:36 AM
To: Brown, James
Cc: simple-evcorr-users@lists.sourceforge.net
Subject: Re: [Simple-evcorr-users] SEC - programmed ignore
Jim,
Below is what I have so far from an earlier discussion with my Tier2
support folks.. I haven't found a clean way of interacting directly with
the database, but given that SEC is a log processor, it makes sense to
keep the configuration channel in the log stream. I plan on
recirculating the logged event at some point back to the NMS, but as an
event type that is not processed again.
I see that John Rouillard offered suggestions as well, and they seem to
coincide with my thinking. I think the difficult part for me will not be
SEC config, but rather the interaction with the change control system in
terms of entering change control information and extracting it from the
tables programatically or via SQL.
Thanks,
- Tim
NMS Change Control process outline
1. Need Oracle or MySql table to communicate changes
ChangeControlTable
In sql parlance:
Create TABLE "ChangeControlTable"
(
"Start" DATE,
"Stop" DATE,
"Duration" NUMBER(10,0),
"HostPortOrPeer" VARCHAR2(30),
"Ticket" VARCHAR(8),
"Complete" BIT
)
2. Need process to extract change items from change control table
Select Start,Stop,Duration,Host,Ticket,Complete
From ChangeControlTable
Where (Start - now < 5 minute) and (Start > 0)
3. Need process to write an event to a monitored syslog log stream.
logger -p local1.notice "$Host change control window $Duration ticket
$Ticket"
4. Need process to extract completed change items from change control table
Select Host,Ticket
From ChangeControlTable
Where (Stop - now > 0) and (Complete = 1)
5. Need process to write the completed event to a monitored syslog stream
logger -p local1.notice "$Host change $Ticket complete"
6. Need Event Correlator rules to read change control events to create
or delete contexts.
# Create a context quiet:<host> with a lifetime of <duration> seconds.
The context
# will be automatically harvested by the correlator. In addition, log
the action taken.
type=Single
ptype=regexp
pattern=(\S+) change control window (\d+) ticket (\S+)
desc=set change control window
context= !quiet:$1
action=create quiet:$1 $2 ; \
logonly %s
# Delete a context quiet:<host> where change is complete
type=Single
ptype=regexp
pattern=(\S+) change complete ticket (\S+)
desc=set change control window
action=create quiet:$1 $2
7. Need to modify any of the existing ticket methods to account for the
new context labeled 'quiet'.
type=PairWithWindow
ptype=RegExp
pattern=Node Unreachable: (\S+)
desc=Node Unreachable:down message
context= !quiet:$1 && !reboot:$1
action=create "reboot:$1" 1200 ; logonly ticket requested %s ;\
spawn SCTicket \
-category "DATA SERVICE" \
-subcategory "data switch" \
-problemtype "connectivity" \
-producttype "none" \
-severity major \
-title "%s" \
-text "The host $1 is not responding to polls. Please investigate."
ptype2=RegExp
pattern2=Node Reachable: (\S+)
desc2=Node Reachable:Up message
action2=delete reboot:$1 ; logonly "Node $1 Reachable - possible bounce"
window=300
--
Tim Peiffer
Network Support Engineer
Office of Information Technology
University of Minnesota/NorthernLights GigaPOP
+1 612 626-7884 (desk)
Note: The information contained in this message may be privileged and
confidential and protected from disclosure. If the reader of this message is
not the intended recipient, or an employee or agent responsible for delivering
this message to the intended recipient, you are hereby notified that any
dissemination, distribution or copying of this communication is strictly
prohibited. If you have received this communication in error, please notify us
immediately by replying to the message and deleting it from your computer.
Thank you. ThruPoint, Inc.
------------------------------------------------------------------------------
ThinkGeek and WIRED's GeekDad team up for the Ultimate
GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the
lucky parental unit. See the prize list and enter to win:
http://p.sf.net/sfu/thinkgeek-promo
_______________________________________________
Simple-evcorr-users mailing list
Simple-evcorr-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users
