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 "Collaboration Tools Availability and Continuity Plan"

From EGIWiki
Jump to navigation Jump to search
(38 intermediate revisions by 2 users not shown)
Line 2: Line 2:
[[Category:Operations]]
[[Category:Operations]]
Back to main page: [[Services Availability Continuity Plans]]
Back to main page: [[Services Availability Continuity Plans]]
'''work in progress'''
Service Availability and continuty plan structure
- Performances (A/R targets agreed in the OLA, comment about the past behaviour and any particular issue)
- summary of the risks assessment
- Av.Co. tests and results


= Introduction =
= Introduction =
Line 25: Line 14:
|-
|-
! scope="row"| Risks assessment
! scope="row"| Risks assessment
| 2018-04-24
| 2019-12-02
| 2019 April
| 2020 December
|-
|-
! scope="row"| Av/Co plan and test
! scope="row"| Av/Co plan and test
| in progress
| 2018-10-26
| --
| --
|-
|-
|}
|}
Previous plans are collected here: https://documents.egi.eu/secure/ShowDocument?docid=3541


= Performances =
= Performances =
The performances reports in terms of Availability and Reliability are produced by [http://argo.egi.eu/ar-opsmon ARGO] on an almost real time basis and they are also collected every 6 months into the Documentation Database (last period [https://documents.egi.eu/document/3270 July 2017 - December 2017].


In the OLA it was agreed the following performances targets, on a monthly basis:
In the OLA it was agreed the following performances targets, on a monthly basis:
Line 42: Line 31:
*Reliability 99%
*Reliability 99%


Over the past years, the Collaboration tools hadn't particular Av/Co issues highlighted by the performances that need to be further investigated.
Other availability requirements:
*the service is accessible either through X509 certificate or via username/password or via EGI Check-in
*the service is accessible via webUI
 
The service availability is regularly tested by nagios probe org.nagiosexchange.Portal-WebCheck and org.nagiosexchange.RT-WebCheck: https://argo-mon.egi.eu/nagios/cgi-bin/status.cgi?servicegroup=SITE_GRIDOPS-CTOOLS_egi.Portal&style=overview https://argo-mon.egi.eu/nagios/cgi-bin/status.cgi?servicegroup=SITE_GRIDOPS-CTOOLS_eu.egi.rt&style=overview
 
The performances reports in terms of Availability and Reliability are produced by [http://egi.ui.argo.grnet.gr/egi/OPS-MONITOR-Critical ARGO] on an almost real time basis and they are also periodically collected into the [https://documents.egi.eu/public/ShowDocument?docid=2324 Documentation Database].
 
Some issues occurred to the Collaboration tools have been notified and investigated, and a plan to improve the service quality is proceeding overseen by EGI Operations.


= Risks assessment and management =
= Risks assessment and management =


For more details, please look at the [https://docs.google.com/spreadsheets/d/1KWfMyLv51BPg-XFCK5zHXuIrLbmKfFBBSIqV4Yn25P4/edit#gid=1565906040 google spreadsheet]. We will report here a summary of the assessment.
For more details, please look at the [https://docs.google.com/spreadsheets/d/1KWfMyLv51BPg-XFCK5zHXuIrLbmKfFBBSIqV4Yn25P4/edit#gid=1939232330 google spreadsheet]. We will report here a summary of the assessment.


== Risks analysis ==
== Risks analysis ==
Line 68: Line 65:
|-
|-
| 2
| 2
| Service unavailable / loss of data due to issues with database server (aldor3)
| DocDB and Confluence, remaining services are using it's own database
| virtualization on HA platform, backups, removing dependency if not necessary
| style="background: yellow"| Medium
| up to 8 hours (1 working day)
| Now DocumentDB, pakiti and rt are the only services that rely on this DB, so the impact of the risk has been decreased to "moderate", and the risk level is now "medium"
|-
| 3
| Service unavailable / loss of data due to software failure
| Service unavailable / loss of data due to software failure
| depends on affected sw, probably all services
| all services depends on quality of open source software, but there's no global dependecy on one software compoment (besides linux kernel and distribution)
| monitoring of system health, backups
| monitoring of system health, backups
| style="background: yellow"| Medium
| style="background: yellow"| Medium
Line 75: Line 80:
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 3
| 4
| service unavailable / loss of data due to human error
| service unavailable / loss of data due to human error
| depends on affected sw/data, probably all services
| depends on affected sw/data, probably all services
| monitoring of system health, backups, actively maintained documentation (wiki)
| monitoring of system health, backups, actively maintained documentation (wiki), subscirption to support forums and chat via IRC (software Indico)
| style="background: yellow"| Medium  
| style="background: yellow"| Medium  
| up to 8 hours (1 working day)
| up to 8 hours (1 working day)
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 4
| 5
| service unavailable for network failure (Network outage with causes external of the site)
| service unavailable for network failure (Network outage with causes external of the site)
| all services (could affect only selected users, depends on problematic networks)
| all services (could affect only selected users, depends on problematic networks)
| monitoring of service availability, alternative network routes
| monitoring of service availability, alternative network routes
| style="background: green"| Low
| style="background: yellow"| Medium
| up to 4 hours (half working day)
| up to 4 hours (half working day)
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 5
| 6
| Unavailability of key technical and support staff (holidays period, sickness, ...)
| Not enough people for maintaining and operating the service
| depends on the problem requiring staff attention, could escalate to all services
| depends on the problem requiring staff attention, could escalate to all services
| contacts to other local staff capable of administering the services
| contacts to other local staff capable of administering the services
Line 99: Line 104:
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 6
| 7
| Major disruption in the data centre. Fire, flood  or electric failure for example
| Major disruption in the data centre.
| all services
| all services
| access to other data centres, geographical diverse backups
| access to other data centres, geographical diverse backups
Line 107: Line 112:
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 7
| 8
| Major security incident. The system is compromised by external attackers and needs to be reinstalled and restored.
| Major security incident. The system is compromised by external attackers and needs to be reinstalled and restored.
| depends on compromised subsystem and/or data, could escalate to all services
| all services
| monitoring of system health, security audits, backups, following best practices for security configuration and timely implementation of patches
| improved monitoring of installed versions, monitoring of network traffic by NREN
| style="background: yellow"| Medium  
| style="background: yellow"| Medium  
| up to 8 hours (1 working day)
| up to 8 hours (1 working day)
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
|-
|-
| 8
| 9
| (D)DOS attack. The service is unavailable because of a coordinated DDOS.
| (D)DOS attack. The service is unavailable because of a coordinated DDOS.
| depends on the attack, could extend from one specific service to complete network outage (=all services)
| all services
| monitoring of service availability, alternative network routes, antiDDoS measures at ISP network
| leveraging of HA platform features, support of network provider
| style="background: green"| Low
| style="background: yellow"| Medium
| up to 4 hours (half working day)
| up to 4 hours (half working day)
| the measures already in place are considered satisfactory and risk level is acceptable
| the measures already in place are considered satisfactory and risk level is acceptable
Line 126: Line 131:
== Outcome ==
== Outcome ==


The level of all the identified risks is acceptable and the countermeasures already adopted are considered satisfactory
There are plans to reduce the level of risk 2) by removing the dependency on aldor3 DB from Confluence and Document Database; moreover it is ongoing a separation process of the several tools by moving them on different machines in order to improve the availability and continuity of the whole service. EGI Operations is following closely the service status with monthly meetings with CESNET team.
 
'''Update 2020-08-07''': the separation of the several tools has been completed so risk 2 has been decreased to "medium".
 
== Additional information ==
 
*procedures for the several countermeasures to invoke in case of risk occurrence are available to the provider
*the Availability targets don't change in case the plan is invoked.
*recovery requirements:
**Maximum tolerable period of disruption (MTPoD) (the maximum amount of time that a service can be unavailable or undelivered after an event that causes disruption to operations, before its stakeholders perceive unacceptable consequences): 1 day
**Recovery time objective (RTO) (the acceptable amount of time to restore the service in order to avoid unacceptable consequences associated with a break in continuity): 1 day
**Recovery point objective (RPO) (the acceptable latency of data that will not be recovered): 1 day
*approach for the return to normal working conditions as reported in the risk assessment.


= Availability and Continuity test =
= Availability and Continuity test =
'''Currently the several components of the service are being migrated and reinstalled to implement a "service separation" with the aim to reduce the incidents propagation from one component to another and improve the overall quality of the service. At the moment we believe a new recovery/continuity test is not necessary, the previous one is still considered to be valid.'''
== Last Availability and Continuity test ==


The proposed A/C test will focus on a recovery scenario: the service has been disrupted and needs to be reinstalled from scratch.
The proposed A/C test will focus on a recovery scenario: the service has been disrupted and needs to be reinstalled from scratch.
The time spent for restoring the service will be measured, using the last backup of the data stored in it (how much information would be lost?)(risks 1,2, and 6?)(how many other operational tool would be affected?).
The time spent for restoring the service will be measured, using the last backup of the data stored in it.
Performing this test will be useful to spot any issue in the recovery procedures of the service.


Performing this test will be useful to spot any issue in the recovery procedures of the service.
'''Test details''':
*The recovery process has been tested and it took 26 minutes.
*Backups are created every two days, so in the worst scenario we can lose data two days back.
 
'''Outcomes and recommendations''':
The test on the whole can be considered successful: the recovery time is acceptable, even though we need to evaluate if loosing 2 days data in the worst case can be tolerable. For some services included in the Collaboration Tools we might need an higher backups frequency: we are going to perform a Business Impact Analysis for these services and then we will agree with the providers the necessary updates to the plan.


= Revision History  =
= Revision History  =
Line 148: Line 175:
| 2018-04-25
| 2018-04-25
| first draft, discussing with the provider
| first draft, discussing with the provider
|-
| <br>
| Alessandro Paolini
| 2018-10-26
| recovery test performed, plan finalised
|-
| <br>
| Alessandro Paolini
| 2019-11-25
| starting the yearly review....
|-
| <br>
| Alessandro Paolini
| 2019-12-05
| wiki page finalised
|-
| <br>
| Alessandro Paolini
| 2020-08-07
| several servers have now their own DB and do not depend any more on aldor3: level of risk 2 decreased to medium
|-
|-
|
|

Revision as of 14:34, 7 August 2020

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


Back to main page: Services Availability Continuity Plans

Introduction

This page reports on the Availability and Continuity Plan for the EGI Collaboration tools and it is the result of the risks assessment conducted for this service: a series of risks and treats has been identified and analysed, along with the correspondent countermeasures currently in place. Whenever a countermeasure is not considered satisfactory for either avoiding or reducing the likelihood of the occurrence of a risk, or its impact, it is agreed with the service provider a new treatment for improving the availability and continuity of the service. The process is concluded with an availability and continuity test.

Last Next
Risks assessment 2019-12-02 2020 December
Av/Co plan and test 2018-10-26 --

Previous plans are collected here: https://documents.egi.eu/secure/ShowDocument?docid=3541

Performances

In the OLA it was agreed the following performances targets, on a monthly basis:

  • Availability: DNS 99%; other services 95%
  • Reliability 99%

Other availability requirements:

  • the service is accessible either through X509 certificate or via username/password or via EGI Check-in
  • the service is accessible via webUI

The service availability is regularly tested by nagios probe org.nagiosexchange.Portal-WebCheck and org.nagiosexchange.RT-WebCheck: https://argo-mon.egi.eu/nagios/cgi-bin/status.cgi?servicegroup=SITE_GRIDOPS-CTOOLS_egi.Portal&style=overview https://argo-mon.egi.eu/nagios/cgi-bin/status.cgi?servicegroup=SITE_GRIDOPS-CTOOLS_eu.egi.rt&style=overview

The performances reports in terms of Availability and Reliability are produced by ARGO on an almost real time basis and they are also periodically collected into the Documentation Database.

Some issues occurred to the Collaboration tools have been notified and investigated, and a plan to improve the service quality is proceeding overseen by EGI Operations.

Risks assessment and management

For more details, please look at the google spreadsheet. We will report here a summary of the assessment.

Risks analysis

Risk id Risk description Affected components Established measures Risk level Expected duration of downtime / time for recovery Comment
1 Service unavailable / loss of data due to hardware failure all services virtualization on HA platform, backups Medium 1 or more working day the measures already in place are considered satisfactory and risk level is acceptable
2 Service unavailable / loss of data due to issues with database server (aldor3) DocDB and Confluence, remaining services are using it's own database virtualization on HA platform, backups, removing dependency if not necessary Medium up to 8 hours (1 working day) Now DocumentDB, pakiti and rt are the only services that rely on this DB, so the impact of the risk has been decreased to "moderate", and the risk level is now "medium"
3 Service unavailable / loss of data due to software failure all services depends on quality of open source software, but there's no global dependecy on one software compoment (besides linux kernel and distribution) monitoring of system health, backups Medium up to 8 hours (1 working day) the measures already in place are considered satisfactory and risk level is acceptable
4 service unavailable / loss of data due to human error depends on affected sw/data, probably all services monitoring of system health, backups, actively maintained documentation (wiki), subscirption to support forums and chat via IRC (software Indico) Medium up to 8 hours (1 working day) the measures already in place are considered satisfactory and risk level is acceptable
5 service unavailable for network failure (Network outage with causes external of the site) all services (could affect only selected users, depends on problematic networks) monitoring of service availability, alternative network routes Medium up to 4 hours (half working day) the measures already in place are considered satisfactory and risk level is acceptable
6 Not enough people for maintaining and operating the service depends on the problem requiring staff attention, could escalate to all services contacts to other local staff capable of administering the services Medium 1 or more working day the measures already in place are considered satisfactory and risk level is acceptable
7 Major disruption in the data centre. all services access to other data centres, geographical diverse backups Medium 1 or more working day the measures already in place are considered satisfactory and risk level is acceptable
8 Major security incident. The system is compromised by external attackers and needs to be reinstalled and restored. all services improved monitoring of installed versions, monitoring of network traffic by NREN Medium up to 8 hours (1 working day) the measures already in place are considered satisfactory and risk level is acceptable
9 (D)DOS attack. The service is unavailable because of a coordinated DDOS. all services leveraging of HA platform features, support of network provider Medium up to 4 hours (half working day) the measures already in place are considered satisfactory and risk level is acceptable

Outcome

There are plans to reduce the level of risk 2) by removing the dependency on aldor3 DB from Confluence and Document Database; moreover it is ongoing a separation process of the several tools by moving them on different machines in order to improve the availability and continuity of the whole service. EGI Operations is following closely the service status with monthly meetings with CESNET team.

Update 2020-08-07: the separation of the several tools has been completed so risk 2 has been decreased to "medium".

Additional information

  • procedures for the several countermeasures to invoke in case of risk occurrence are available to the provider
  • the Availability targets don't change in case the plan is invoked.
  • recovery requirements:
    • Maximum tolerable period of disruption (MTPoD) (the maximum amount of time that a service can be unavailable or undelivered after an event that causes disruption to operations, before its stakeholders perceive unacceptable consequences): 1 day
    • Recovery time objective (RTO) (the acceptable amount of time to restore the service in order to avoid unacceptable consequences associated with a break in continuity): 1 day
    • Recovery point objective (RPO) (the acceptable latency of data that will not be recovered): 1 day
  • approach for the return to normal working conditions as reported in the risk assessment.

Availability and Continuity test

Currently the several components of the service are being migrated and reinstalled to implement a "service separation" with the aim to reduce the incidents propagation from one component to another and improve the overall quality of the service. At the moment we believe a new recovery/continuity test is not necessary, the previous one is still considered to be valid.

Last Availability and Continuity test

The proposed A/C test will focus on a recovery scenario: the service has been disrupted and needs to be reinstalled from scratch. The time spent for restoring the service will be measured, using the last backup of the data stored in it. Performing this test will be useful to spot any issue in the recovery procedures of the service.

Test details:

  • The recovery process has been tested and it took 26 minutes.
  • Backups are created every two days, so in the worst scenario we can lose data two days back.

Outcomes and recommendations: The test on the whole can be considered successful: the recovery time is acceptable, even though we need to evaluate if loosing 2 days data in the worst case can be tolerable. For some services included in the Collaboration Tools we might need an higher backups frequency: we are going to perform a Business Impact Analysis for these services and then we will agree with the providers the necessary updates to the plan.

Revision History

Version Authors Date Comments

Alessandro Paolini 2018-04-25 first draft, discussing with the provider

Alessandro Paolini 2018-10-26 recovery test performed, plan finalised

Alessandro Paolini 2019-11-25 starting the yearly review....

Alessandro Paolini 2019-12-05 wiki page finalised

Alessandro Paolini 2020-08-07 several servers have now their own DB and do not depend any more on aldor3: level of risk 2 decreased to medium