Maintenance windows
Table of Contents |
---|
TDS team reserves regular maintenance window maintenance windows are scheduled every Wednesday between 5-7 PM CET(winter time) or CEST (summer time). During that time frame planned changes are deployed to production environments. Service outages are usually 10-15 minutes, maximum 30 minutes. In case of expected longer outage all relevant users would be informed in advance by relevant channels.
All planned updates are communicated to users in advance. Following information channels are utilized for notifications about upcoming updates:
. Operations team deploys service upgrades, patches, or changes. Most deployments are implemented within 15 minutes, however more complex changes can take 30 minutes or longer. Users are informed about all planned changes in advance, longer outages are highlighted in the communication.
Planned Changes
Planned changes deliver regular service upgrades and infrastructure changes.
Service | Communication Channel and Timeframe |
---|---|
Jira and Confluence | Application banner - 48h in advance |
...
Artifactory, Bitbucket, Gitlab, Subversion, SeedDMS |
...
Application banner - 48h in advance |
...
(not available for SeedDMS and Subversion) |
...
E-mail - 5 |
...
calendar days in advance, addressed to TDS project owners and SaaS project |
...
/space/repository admins (Users for Subversion) |
...
* | |
Dedicated SaaS (SaaS applications dedicated to one or multiple TDS projects) | SPOC and named stakeholders are informed by agreed channels 5 calendar days in advance Application banner 48 hours prior to the change (if requested by SPOC) |
Emergency changes
TDS team reserves the right to take needed urgent actions to deploy changes, patches, workaround and other corrective actions in case of critical security issues to prevent security incidents. Key stakeholders (or SPOC) are informed by agreed channels about ongoing activities. During such emergency activities some services might be affected by temporary outages. The TDS team communicates the emergency changes to the Customer as soon as possible (banner, mail, or chat).
- Servers - each project is always responsible for its own servers, however TDS team is usually performing security patching to mitigate critical security issues without delays to avoid any incidents
- (unless customer implemented the security patch within given timeframe)
- in case of expected downtime project owners/admins are informed by email (or chat platform when applicable)
- key stakeholders (or SPOC) are informed by agreed channels
TDS Outage and Upgrade Communication
TDS maintenance windows are scheduled every Wednesday 5-7 PM CET. Operations team deploys service upgrades, patches, or changes. Typical service outage takes less than 15 minutes. More complex changes can take more than 30m and this is highlighted in user communication.
Communication Scheme
All production changes are announced at least 5 business days in advance.
Service | Communication in Tietoevry | Communication in External Customer Environment |
---|---|---|
Jira and Confluence | Application banner | Application banner Named stakeholders including SPOC are informed according to agreed channels. |
Artifactory, Bitbucket, Gitlab, Subversion, SeedDMS | Application banner Relevant users are informed by email (This is not clear and needs to be more specific.) | Application banner Named stakeholders including SPOC are informed according to agreed channels. |
Services dedicated to TDS Project | Application banner All project users are informed via e-mail | Application banner (if preferred by SPOC) Named stakeholders including SPOC are informed according to agreed channels. |
Servers | Project owners and server admins are informed by email (or chat platform when applicable) | Named stakeholders (or SPOC) are informed by agreed channels. |
TDS Portal | TDS portal is upgraded frequently through the automated continuous delivery pipeline with a few seconds of downtime in most cases. Since this is not a business-critical application and upgrades are performed very frequently, users are not informed about outages. New functionality is described in Release Notes. |
*) Deployed during 2022-10 – respective toolkit needs to be implemented to generate list or e-mail recipients
Regular schedule of maintenance windows Due to TDS operational reasons, regular schedule can be occasionally altered (e.g. longer outage outages are scheduled for the weekend instead of Wednesday evening). All irregular changes are announced according to Communication Scheme above.
Emergency Changes
Emergency changes (typically urgent high or critical security patches) are deployed as soon as possible, preferably after 5 PM regardless of maintenance windows. Users are informed according to Communication Scheme above, however through the same communication channels as for planned changes, but the time between user communication and service outage/ change deployment is less than 5 daysshorter. Note that critical security vulnerabilities are often patched within 24 hours after CVE is published.
Servers
Virtual servers are dedicated to TDS projects and project teams are responsible for performing regular server maintenance. In case of critical security vulnerabilities, the TDS operations team patches these servers with high priority as soon as possible. Automatic security updates are enabled by default on CentOS and Ubuntu servers, hence intervention from TDS operations team should be exceptional. Server admins (and named SPOC if applicable) are informed in advance as soon as possible via e-mail or chat.
CVE's
To check current CVE mitigation rules please navigate to CVE Mitigations
Related Pages
Roadmap (best-effort development team plan) | Release Notes | Service description (Including Service Level Agreement) | Service Status | TDB - ?Roadmap