Participants ============ 1. guilhem 2. Brett
Agenda ====== * New infra status page (https://status.documentfoundation.org , powered by cachetHQ) + https://redmine.documentfoundation.org/issues/1079#note-20 + Meant to give users a quick overview (HTTP status and response time only), not to replace the more comprehensive internal infra metrics + Atom & RSS feeds, email subscription → best not to subscribe using a TDF-hosted address! + 24 HTTP sites monitored at the moment; more can be added later + Metrics and outages are automatically filled using the cachet API and https://github.com/CastawayLabs/cachet-monitor; scheduled maintenance can be added too, and text/status can be edited manually to inform the community about our progress in solving an issue + Current limitations: - no v4/v6 distinction; ideally the component state would change to “outage” if IPv6 (resp. v4) connectivity is broken, assuming there is an AAAA (resp. A) record - while traffic is not routed privately, DNS queries are; better to use upstream DNS servers instead, and use gustl for .tdf only → solutions: use systemd-{networkd,resolved}, or a local caching recursive resolver - would be nice to get SMTP & IMAP status codes, too + TODO: add a CNAME + 301 redirect for status.libreoffice.org + TODO: subscribe IZBot to the feed. Can cachet push updates to it? Brett: don't think so + Officially announce the page at LibOCon, but folks can subscribe already * Prometheus alert system (Brett) + Merged and deployed; only email notifications for now (no SMS) * Prometheus: downgraded node-exporter from stretch-backports to stretch on the monitoring box * Upgrades: + Question: status of tb31.libreoffice.org? (last Ubuntu — 14.04.5 LTS — box, would like to align on our current — Debian 9.4 — baseline instead) + 37 boxes still on Debian Jessie (24 prod boxes incl. 3 hypervisors), need to be updated before the end of year ideally - some of these require more work (not based on the Jessie baseline), eg dashboard (vm167) - tentative plan: migrate the dashboard to vm213 (fresh Stretch VM) during LibOCon + upgraded gerrit to stretch last week (got rid of salt states gerrit vs. gerrit.prod), improved systemd unit files * Single Sign On + Auth method switched from SilverStripe's own backend to SAML 2.0 on the LibreOffice & TDF websites (SS3/newdesign) + Working on the migration for the blog's (WordPress) auth backend now * Next call: Tuesday October 16 2018 at 18:30 Berlin time (16:30 UTC). -- Guilhem. -- To unsubscribe e-mail to: [email protected] Problems? https://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/ Posting guidelines + more: https://wiki.documentfoundation.org/Netiquette List archive: https://listarchives.libreoffice.org/global/website/ Privacy Policy: https://www.documentfoundation.org/privacy
