Disaster Recovery Documentation for Polaris
I.
Personnel
a.
DBA Team
i.
Scott Dove, cell 214.486.5572
ii.
Shane Sikora, cell 817.480.3242
b.
Application Team
i.
Don Sikes, cell 214.507.5893, home 817.477.5051
ii.
Kassie Rangal, cell
817.690.6804
c.
SE AIX Team
i.
Mark Jackson, office 214-879-1634 cell 214.502.0244
ii.
Joe Stephens, office 214.879.1638 cell 214-202-7104
d.
Storage Team
i.
II.
Assumptions
a.
Oracle and the pol2p database is ready on
TDC server.
b.
File directory exists exactly as it does currently on the
mxfmtxbsdm01 AIX server.
c.
Users are configured with permission's as they are on the BSDM01
server.
d.
All files have been mirrored to exist on the TDC server.
e.
TDC server name will NOT be different than current server and will
resume the identity of the current server.
f.
Host names are used by internal jobs on both the mainframe and
GISB servers that forward jobs to MXFBSDM01.
g.
Parser should come up based on the cloning of the system.
III.
Overview of the Testing
of the Recovery Plan
a.
Database has been cloned from the last hot-backup of production.
b.
Database has been brought up on the DR server and verified
operational.
c.
Parser will not be tested due to identified conflicts with having
the production Parser operational.
d.
The only difference between the DR site and the current production
site will be the IP addresses of the servers.
e.
Operating system was built by cloning everything in the root
volume group.
f.
User accounts and passwords were cloned.
g.
CRON schedules, security, system parameters were also cloned.
IV.
Recovery Plans
a.
Rename / re-point DNS name from MXDAPPLS01 to MXFBSDM01. Done
by the
b.
Restore backup DB files from TSM. Done by the DBA Team and Storage
Team. This will recover the data to a 7 day recovery point. Additional steps
(documented below) are required to pull the RPO to within 24 hours.
c.
Start DB abd issue media recovery from
logs. Done by the DBA Team.
d.
Begin application checkout. Done by the Application Team.
e.
Manual load files from ERCOT and EPN. Done by the DBA and
application Teams. This applies all transactions to current day and meets the
24 hour RPO objective.
f.
Uncomment CRON jobs. This is done by the DBA and
g.
Update all data feed configuration files as well as code(where necessary) to reflect the new TDC server. Done by
the Application Team. This will be used by all feeds that ftp us information
hourly/daily. The feeds are listed in order of importance. All these data feeds
currently use the etcamftp or jparser
user id.
i.
Data Feeds
1.
Gateway Jobs - Confirm Properties Files
2.
CIS Jobs – Confirm the following:
PCDTHADD PCDTHINS PCDTHKLC
PCDTHKLI PCDTHKLX PCDTHMTR
PCDTHOPC PCDTHOPI PCDTHOPR
PCDTHOPX PCDTHPRE PCDTHSOR
PCDTH100 PCDTH200 PCDTH201
PCDTH202 PCDTH203 PCDTH204
PCDTH205 PCDTH206 PCDTH207
PCDTH280 PCDTH300 PCDTH301
PCDTH302 PCDTH970 PCDTH997
PCISSIDB PCSO5012 PCSO5013
3.
ERCOT Jobs - Confirm Properties Files
4.
Mellon Jobs - Change data feed code to reflect new server name.
h.
Confirm all Database connections within the Application (parser/web).
Done by the Application Team.
i.
The web portion (the part currently running on the Linux box) of
the application will need a Properties File Change.
ii.
The remaining application will need to have code changed to
reflect the new server. This refers to those files that have the server name
hard-coded within them.
i.
Ensure that ip address/dns for web portal is configured for TDC server. Done by
the
j.
Launch Tomcat for web portion of Polaris. Done by the Application
Team.
k. Polaris should be running and functional. Done by the Application Team.