Differences

This shows you the differences between two versions of the page.

--- infra:hscloud [2025/05/18 21:22] – radex
+++ infra:hscloud [2026/06/05 06:15] (current) – radex
@@ Line 9: / Line 9: @@
 ===== Components =====
-Currently hscloud is made up of a Kubernetes cluster named k0.hswaw.net. It runs on the following machines: bc01n01, bc01n02, bc01n03, dcr01s22, dcr01s24. In total we have 328GB of RAM and 144 x86 cores. We also have a half PB of storage on old SAS drives (most of it currently cold), accessible via Ceph (radosgw or Kubernets PersistentVolumes).
+Currently hscloud is made up of two Kubernetes clusters named:
+  * k0.hswaw.net - it runs on the following machines: dcr01s22, dcr01s24, dcr03s16. It has 144 x86 cores and 368GB of RAM.
+  * k1.hswaw.net (new, wip, more details later). It runs on the following machines: dcr03s19n01, dcr03s19n02, dcr03s19n03. It has 144C/288T in x86 cores (EPYC 7451), and 386 GB of RAM.
+We also have ~200TB of storage on old SAS drives + ~6TB of fast NVMe SSD storage, accessible via Ceph (radosgw or Kubernetes PersistentVolumes) across two Ceph clusters: waw3 (old, main, currently ~100TB), waw4 (new, wip, currently ~70TB).
 ===== Tenants =====
@@ Line 21: / Line 26: @@
 Here's a list of services that currently live on Boston Packets, but we'd like to migrate to hscloud. Ask on #infra on how to contribute.
-  * [[https://owncloud.hackerspace.pl]] - work in progress.
   * mailman ([[https://lists.hackerspace.pl]]) - the web service is already on k8s, but also proxied via boston. Mailman-core and the database (used by both web and core, postgres) are still on boston. (2025-01: on implr's todo list)
-  * ldap/kerberos - (hard) (2025-01: infowski has a WIP ldap replica)
+  * ldap/kerberos - (hard) (WIP, we already have a replica outside boston)
   * email services (exim, dovecot) - (hard)
   * mun (irc bot)
-  * ood (irc bot)
+  * `~user` dir serving
+  * sensitivefilter.py
 ===== k0 Evacuation Aktion =====
@@ Line 34: / Line 39: @@
 <code>
 ## Ready to migrate
-personal-*
-## Waiting for k1 ceph
-factorio
-inventory
-minecraft
 personal-*
-sourcegraph
-valheim
-## Stage 2 - sensitive / important
-capacifier
-internet
 mailman-hackerspace-prod
 mastodon-hackerspace-{prod,qa}
-nextcloud
+codehosting-prod            [ prepared, copy data and switch over when ready ]
-paperless
-roundcube
-speedtest
-walne
-wiki
-## Stage 3: Kritische infrastruktur
-codehosting-prod
 gerrit
-kasownik
-ldapweb
 matrix{,-0x3c}
-redmine
+kasownik                    [ prepared, copy data and switch over when ready ]
-registry
+monitoring-global
-site
-sso
 ## Already migrated
@@ Line 85: / Line 63: @@
 printservant
 zhp-site
+internet
+speedtest
+walne
+ood
+sourcegraph
+inventory
+wiki
+paperless
+redmine
+site
+capacifier
+ldapweb
+sso
+registry
+roundcube
+asterisk
+nextcloud
 </code>
-===== Planned upgrades =====
+List of unfinished business before we can make k1 production-grade and retire k0:
+  * Storage: Migrate all ceph rbd data
+  * Storage: Migrate ceph rgw
+  * Storage: Add benji
+  * Migrate remaining services
+  * Add global/default NetworkPolicies
+Also todo to improve k1 reliability:
-See [[projects:crashcloud|crashcloud]].
+  * more reliable LoadBalancer routing (fuckin' calico maaan)
 ===== Monitoring =====
@@ Line 102: / Line 105: @@
 See [[https://hackdoc.hackerspace.pl/doc/codelabs/cookbook.md|Kubernetes Cookbook]]
+===== dcr03s19 =====
+This machine is a [[https://www.supermicro.com/en/Aplus/system/2U/2123/AS-2123BT-HTR.cfm|Supermicro A+ Server 2123BT-HTR]], made up of 4 [[https://www.supermicro.com/en/products/motherboard/H11DST-B|H11DST-B]] nodes. Each has 2x 24-core 48-thread AMD EPYC CPUs.
+Nodes a,b,c (dcr03s19n{01,02,03}) make up the k1 cluster. The final node is Snowflake, a VM machine (hosting, among other things, Boston Packets).
+Manuals:
+{{infra:mnl-2069.pdf}}
+{{infra:mnl-2051.pdf}}