Alert.png The wiki is deprecated and due to be decommissioned by the end of September 2022.
The content is being migrated to other supports, new updates will be ignored and lost.
If needed you can get in touch with EGI SDIS team using operations @ egi.eu.

Difference between revisions of "PROC16 Decommissioning of unsupported software"

From EGIWiki
Jump to navigation Jump to search
Line 15: Line 15:
= Overview  =
= Overview  =


The Service type decommission procedure was created to define steps which have to be taken to remove service type from prodcution infrastructure. <br>
The Service type decommission procedure was created to define steps which have to be taken to remove service type from prodcution infrastructure. <br>  


= Terminology  =
= Terminology  =


'''Support deadline'''
'''Support deadline'''  
*from this day service type is unsupported
 
*from this day service type is unsupported  
*sites '''should not''' register retired services in production infrastructure  
*sites '''should not''' register retired services in production infrastructure  
*NGIs and sites should start action to move their services to supported versions
*NGIs and sites should start action to move their services to supported versions


'''Decommission deadline'''
'''Decommission deadline'''  
*from this day '''no services for the retired service type can be present in production infrastructure'''
 
*in case of violation security team can remove the site from production infrastructure
*from this day '''no services for the retired service type can be present in production infrastructure'''  
*in case of violation security team can remove the site from production infrastructure  
*service type is removed from operations tools
*service type is removed from operations tools


= Steps  =
= Steps  =


Following steps are taken since '''support deadline''' is known for COO
Following steps are taken since '''support deadline''' is known for COO  
 
#COO on OMB announce '''decommission deadline''' for service migration
#The broadcast is send to NGI managers and Site administrators
#NGI managers announce the information to the sites
#Documentation about decommissioned service should re-classified as obsoleted (I would not remove it from our wiki, but I would just reclassify it)
#New probe is developed for the MW nagios that extracts hostnames from GOCDB and BDII associated to type and raises critical alarm in the operations dashboard
#* ROD teams followup the service migration till the decommission deadline - site admins should provide migration or decommission plan See instruction


# COO on OMB announce '''decommission deadline''' for service migration
After '''Decommision deadline'''  
# The broadcast is send to NGi managers and Site administrators
# NGI managers announce the information to the sites
# Documentation about service X should re-classified as obsoleted (I would not remove it from our wiki, but I would just reclassify it)
# Security probe is developed for the security nagios that extracts hostnames from GOCDB and BDII associated to type X and raises critical alarm in the security dashboard
#* ROD teams followup the service migration till the decommission deadline - site admins should provide migration or decommission plan


After '''Decommision deadline'''
#Probes for the service are removed from profiles:  
# Probes for the service are removed from profiles:
#*ROC  
#*ROC
#*ROC_OPERATORS  
#*ROC_OPERATORS  
#*ROC_CRITICAL  
#*ROC_CRITICAL  
#*the SAM probes at the earliest convenience are removed from the SAM release  
#*the SAM probes at the earliest convenience are removed from the SAM release  
# CSIRT team followup the migration and if needed suspend sites which didn't migrate  
#CSIRT team followup the migration and if needed suspend sites which didn't migrate  
# If a service specific SU in GGUS exists, the SU should be removed from GGUS  
#If a service specific SU in GGUS exists, the SU should be removed from GGUS  
# The service type should be disabled in GOCDB (i.e. service entries can no more declared to be of service type X), but existing instances associated to type X continue to exist  
#The service type should be disabled in GOCDB (i.e. service entries can no more declared to be of the service type), but existing instances associated to the type continue to exist<br>
# Security probe can be removed (should we do that?)
 
== TODO ==
 
* involve quattor WG
 
unresponsive sites handling process
# NGI will be requested by EGI.eu operations to contact site administrators to provide upgrade plans in the site ticket, DEADLINE MON 12/11
# EGI.eu operations will send a VO broadcast warning VO managers about sites risking suspension
# In case of no reply from site administrators, the NGI operations management (NGI manager or deputy, security officer) will be requested to put the affected service end-point in downtime, DEADLINE MON 19/11
# In case of no action by NGI, EGI CSIRT will suspend the site (after MON 19/11)


I would suggest the following procedure for EGI-CSIRT Security Officer on Duty
== Instuctions for ROD teams ==
(SOOD), handling sites that qualified for suspension:
# Open ticket in RT ask COO to suspend site X (CCing  Site/NGI-Manager, Site/NGI-Security-Contact)
# Site/NGI can give a comment here, if we see suddenly some activity from the site we can offer that the site goes immediately in Downtime.
# within 24 hours COO gives the green/red light for Suspension in the ticket
# After having received an OK from COO, SOOD changes the sites "certification status" to "Suspended"


This is in EGI-CSIRT Critical Vulnerability Operational Procedure: https://documents.egi.eu/secure/RetrieveFile?docid=283&version=9&filename=EGI-CSIRT-Procedure-CriticalSecurity-V8.pdf Skipping the extensive communication part which has for these sites already proven to not work, thus we might want to jump directly to


==


4.5 Carry out Site suspension


Site suspension WILL be carried out by the EGI CSIRT co-ordinator or deputy. Site suspension is carried out by changing the status of the site in the GOCDB to ‘suspended’.
   
==


= Revision history  =
= Revision history  =

Revision as of 12:49, 19 November 2012

Main EGI.eu operations services Support Documentation Tools Activities Performance Technology Catch-all Services Resource Allocation Security


Documentation menu: Home Manuals Procedures Training Other Contact For: VO managers Administrators




Title Service type decommission
Document link https://wiki.egi.eu/wiki/PROC16
Last modified 1.0
Policy Group Acronym OMB
Policy Group Name Operations Management Board
Contact Group operations at mailman.egi.eu
Document Status Draft
Approved Date
Procedure Statement A procedure for removal of service type from production infrastructure.
Owner Owner of procedure


Overview

The Service type decommission procedure was created to define steps which have to be taken to remove service type from prodcution infrastructure.

Terminology

Support deadline

  • from this day service type is unsupported
  • sites should not register retired services in production infrastructure
  • NGIs and sites should start action to move their services to supported versions

Decommission deadline

  • from this day no services for the retired service type can be present in production infrastructure
  • in case of violation security team can remove the site from production infrastructure
  • service type is removed from operations tools

Steps

Following steps are taken since support deadline is known for COO

  1. COO on OMB announce decommission deadline for service migration
  2. The broadcast is send to NGI managers and Site administrators
  3. NGI managers announce the information to the sites
  4. Documentation about decommissioned service should re-classified as obsoleted (I would not remove it from our wiki, but I would just reclassify it)
  5. New probe is developed for the MW nagios that extracts hostnames from GOCDB and BDII associated to type and raises critical alarm in the operations dashboard
    • ROD teams followup the service migration till the decommission deadline - site admins should provide migration or decommission plan See instruction

After Decommision deadline

  1. Probes for the service are removed from profiles:
    • ROC
    • ROC_OPERATORS
    • ROC_CRITICAL
    • the SAM probes at the earliest convenience are removed from the SAM release
  2. CSIRT team followup the migration and if needed suspend sites which didn't migrate
  3. If a service specific SU in GGUS exists, the SU should be removed from GGUS
  4. The service type should be disabled in GOCDB (i.e. service entries can no more declared to be of the service type), but existing instances associated to the type continue to exist

Instuctions for ROD teams

Revision history

Version Authors Date Comments