Restoring Services
From GridWiki
After an unexpected outage of power, the system should be restored in the order described below. This order ensures that most critical services are started first, but also that no problems with dependencies occur.
| Service Class | Service Type | Hostname | Comment |
| Management | OpenVPN | schrepel | Required for access to other hosts in farm |
| Wiki | Switch to off-site server. Contains essential information like node function list and misc documentation items | ||
| Externally visible | VOMS server | kuiken | VOMS server for GIN |
| EUGridPMA web site | keerder (zeis, weikuip) | Web server for EUGridPMA. Both are hosted on the Dom0 'keerder'. Alternate services are on dodo and lama - if the DNS servers have been redirected there, after 8 hours of outage these functions have become less critical. | |
| DutchGrid CRL web site | kaasvat | serves the CRLs on ca.dutchgrid.nl/medium/cacrl.pem to the outside world. This service directly affects all grid sites in EGEE, LCG and the world. Kaasvat is in the secured rack #2. | |
| Site core | Monitoring (Nagios) | spade, riek, eg | Nagios servers; provide clue about what is functional and what is not |
| Install server | stal | Needed to install hosts from scratch | |
| Repository servers | stalkaars-01.farm, stalkaars-03.farm | Provide RPM repositories | |
| LDAP server | teugel, hooimijt, stalkaars-0[13].farm | Many other hosts depend on ldap for user authentication, automounts, etc. | |
| Install server | stal | Needed to install hosts from scratch | |
| NFS server | vlaai, hoeve, schuur | NFS server for home directories, pool accounts homes, experiment software area | |
| Site generic with dependencies | Monitoring (Ganglia) | trog | Ganglia monitoring will only work when the collectors nodes per cluster are up. |
| Xen servers | appelvanger, silo, mesthoop, moestuin, hilde, kaf, kribbe | Host environment for other service machines | |
| Database server | bedstee | Provides databases for DPM storage and local LFC | |
| Site BDII | siteinfo03 | Publish avialable resource in information system; host may be virtual machine | |
| BDII | bdii03, graskaas | Top level BDII | |
| log server | boes | Collecting syslog messages | |
| Storage | DPM head node | tbn18 | Storage interface |
| disk server NL-T1 | hooi-ei-*, hooibroei, hooizolder | DPM disk servers for T1 | |
| other disk servers | garitxako, hooikuil, hooischelf, hooivork, hooikist | DPM disk servers for non-T1 users | |
| Computing | Batch System | silo | Torque/Maui |
| CE | gazon, trekker | Computing Elements | |
| UI, VOBOX | tbn12, bosui, erf, kot | Interface to check grid services | |
| WMS, LB, RB | graszode, grasveld, boszwijn, bosheks, dorsvlegel | Workload management systems, required for local job submission | |
| MON, LFC | klomp, opkamer | Misc grid services | |
| WN | wn-val-*, wn-lui2-*, wn-lui1-*, wn-bull-* | Worker nodes. | |
| Other services | DutchGrid CA | hek | request submission interface, processing new requests |
| web, backup | beerput | serves main dutchgrid site, VL-e PoC and NDPF CVS repository. Also does the rsync based backup of must of the non-farm systems | |
| SVN, SSC | sikkel, rooier | svn services on keerder, just check to see if they are there | |
| public mirror | rijf | located in 2nd Valentine rack! |
