EGI Workload Manager Availability and Continuity Plan
|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
This page reports on the Availability and Continuity Plan for EGI Workload Manager - DIRAC4EGI 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.
|Risks assessment||2019-01-10||2021 Apr|
|Av/Co plan and test||2020-04-14||2021 Apr|
Availability requirements and performances
In the OLA it was agreed the following performances targets, on a monthly basis:
- Availability: 99%
- Reliability 99%
Other availability requirements: - the service is accessible through either X509 certificate or OAuth2 IdP (upcoming with the new release) - The service is accessible via CLI and webUI (link to the probes that checks the authentication and the service in general)
The service availability is regularly tested by nagios probe org.nagiosexchange.Portal-WebCheck: https://argo-mon.egi.eu/nagios/cgi-bin/status.cgi?host=dirac.egi.eu
Over the past years, the Workload Manager hadn't particular Av/Co issues highlighted by the performances that need to be further investigated.
Risks assessment and management
For more details, please look at the google spreadsheet. We will report here a summary of the assessment.
|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 the service components||Database protection with regular backups. Redundant Configuration database. Snapshots of virtual machines hosting DIRAC4EGI services||Medium||1 working day after the hardware failure recovery||the measures already in place are considered satisfactory and risk level is acceptable|
|2||Service unavailable / loss of data due to software failure||All the service components||Backup of the MySQL and Configuration databases||Low||1 working day||the measures already in place are considered satisfactory and risk level is acceptable|
|3||service unavailable / loss of data due to human error||All the service components||Backup of the MySQL and Configuration databases||Low||1 working day||the measures already in place are considered satisfactory and risk level is acceptable|
|4||service unavailable for network failure (Network outage with causes external of the site)||All the service components||Geographically distributed redundant Configuration Service. Redundant failover Request Management Service.||Low||1 hour after the network recovery||the measures already in place are considered satisfactory and risk level is acceptable|
|5||Unavailability of key technical and support staff (holidays period, sickness, ...)||Resources management. User support. Security infrastructure components||Automation of synchronization with BDII, VOMS, GocDB information indices. Automated resource monitoring service||Low||1 or more working days||the measures already in place are considered satisfactory and risk level is acceptable|
|6||Major disruption in the data centre. Fire, flood or electric failure for example||All the service components||Backup of the MySQL and Configuration databases||Low||several weeks||the measures already in place are considered satisfactory and risk level is acceptable|
|7||Major security incident. The system is compromised by external attackers and needs to be reinstalled and restored.||All the service components||Backup of the MySQL and Configuration databases||Low||1 or more working days||the measures already in place are considered satisfactory and risk level is acceptable|
|8||(D)DOS attack. The service is unavailable because of a coordinated DDOS.||All the service components||Limited service queries queues avoiding dangerous overloading of the service components. Automatic service restart after going down due to an overload.||Low||1 hour||the measures already in place are considered satisfactory and risk level is acceptable|
The level of all the identified risks is acceptable and the countermeasures already adopted are considered satisfactory
- There aren't special procedures to invoke in case of risk occurrence but the general administrator guide and generic internal procedures
- 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): 5 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 (this has to be less than MTPoD)): 5 days
- Recovery point objective (RPO) (the acceptable latency of data that will not be recovered): n/a
- approach for the return to normal working conditions as reported in the risk assessment.
- The dedicated GGUS Support Unit will be used to report any incident or service request.
- The providers can contact EGI Operations via ticket or email in case the continuity plan is invoked, or to discuss any change to it.
Availability and Continuity test
The proposed A/C test will focus on a recovery scenario: the service is supposed to have been disrupted and needs to be reinstalled from scratch. Typically this covers the risks 1,2, and 7. The last backup of the data will be used for restoring the service, verifying how much information will be lost, and the time spent will be measured.
Performing this test will be useful to spot any issue in the recovery procedures of the service.
More details available on https://documents.egi.eu/document/3597
|Test case||Simulation||Recover time||Actions||Status|
|Service/Agent crash that can be caused by some transaction failure, software error or human error||Kill the component process||few seconds||The component is restarted automatically by the system monitoring facility||PASS|
|Host failure, for example due to a power cut||reboot dirac4.grid.cyfronet.pl server||few minutes||The host rebooting sequence contains an automatic restart of all configured DIRAC components by using supervisord.||PASS|
|Installed software corruption||reinstall DIRAC software stack from scratch||10 - 15 minutes||Manual intervention: running dirac-install installer tool; verify that all the components properly restart.||PASS|
|Configuration files loss or corruption, for example, due to a hard disk failure.||BackUps of the local configuration files in a database or on another server||few minutes||Replace the lost configuration with a backup copy.||PASS|
|DB corruption and/or crash||recover from dump||5 - 30 minutes||Manual intervention by the CYFRONET database service administrators||PASS|
The test can be considered successful: the service can be restored in few time if hardware, software or database failures occur.
|Alessandro Paolini||2019-01-10||first draft, discussing with the provider|
|Alessandro Paolini||2019-08-27||adding other availability requirements, and additional information for the risk assessment|
|Alessandro Paolini||2019-11-25||page updated with additional availability requirements, and additional information section. Waiting for the recovery test, hopefully to be done in January.|
|Alessandro Paolini||2020-04-14||added details about the recovery test provided by the supplier. Plan finalised.|