User Tools

Site Tools


infra:hscloud

Differences

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

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
infra:hscloud [2025/06/03 20:03] – radexinfra:hscloud [2026/06/05 06:15] (current) – radex
Line 12: Line 12:
  
   * k0.hswaw.net - it runs on the following machines: dcr01s22, dcr01s24, dcr03s16. It has 144 x86 cores and 368GB of RAM.   * 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 256 GB 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 a half PB of storage on old SAS drives (most of it currently cold), accessible via Ceph (radosgw or Kubernets PersistentVolumes) across two Ceph cluster: waw3 (old, main, currently ~120TB), waw4 (new, wip, currently ~24TB).+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 ===== ===== Tenants =====
Line 26: 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. 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)   * 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)   * email services (exim, dovecot) - (hard)
   * mun (irc bot)   * mun (irc bot)
-  * ood (ood's postgres is still to be migrated, ol0ck TODO)+  * `~user` dir serving 
 +  * sensitivefilter.py 
 ===== k0 Evacuation Aktion ===== ===== k0 Evacuation Aktion =====
  
Line 39: Line 40:
 ## Ready to migrate ## Ready to migrate
            
-ood (postgres) 
 personal-* personal-*
-inventory                   [ prepared, copy data and switch over when ready ] 
-sourcegraph                 [ prepared, copy data and switch over when ready ] 
-internet                    [ working replica, switch over when ready ] 
-speedtest                   [ working replica, switch over when ready ] 
-walne                       [ working replica, switch over when ready ] 
- 
-factorio                    [ or abandon? ]      
-minecraft                   [ or abandon? ]      
-valheim                     [ or abandon? ] 
- 
-## Stage 2 - sensitive / important 
- 
-nextcloud                   [ ready to prepare ] 
-paperless                   [ ready to prepare ] 
 mailman-hackerspace-prod     mailman-hackerspace-prod    
 mastodon-hackerspace-{prod,qa} mastodon-hackerspace-{prod,qa}
-capacifier                  [ working replica, switch over when ready ] +codehosting-prod            [ prepared, copy data and switch over when ready ]
-roundcube                   [ prepared, copy data and switch over when ready ] +
-wiki                        [ prepared, copy data and switch over when ready ] +
- +
-## Stage 3: Kritische infrastruktur +
- +
-codehosting-prod            +
 gerrit                       gerrit                      
 matrix{,-0x3c}               matrix{,-0x3c}              
-redmine                      
 kasownik                    [ prepared, copy data and switch over when ready ] kasownik                    [ prepared, copy data and switch over when ready ]
-ldapweb                     [ working replica, switch over when ready ] +monitoring-global
-registry                    [ working replica, switch over when ready ] +
-site                        [ working replica, switch over when ready ] +
-sso                         [ working replica, switch over when ready ]+
  
 ## Already migrated ## Already migrated
Line 87: Line 63:
 printservant                 printservant                
 zhp-site   zhp-site  
 +internet                    
 +speedtest                   
 +walne     
 +ood                 
 +sourcegraph                  
 +inventory                   
 +wiki                        
 +paperless                   
 +redmine                     
 +site                        
 +capacifier                  
 +ldapweb                     
 +sso                         
 +registry                    
 +roundcube
 +asterisk          
 +nextcloud
 </code> </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 ===== ===== Monitoring =====
infra/hscloud.1748981030.txt.gz · Last modified: by radex

Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki