Gina specifications

Gina is the Grid cluster at SURFsara; currently all the servers are located at the Digital Realty datacenter. After different tenders these servers belong to different brands (Dell & Fujitsu). Three service nodes are used to host virtual machines for different purposes (creamces, monitoring, installation, dns servers…). All the worker nodes are installed on physical nodes.

This page describes the specifications for the Gina Grid cluster at SURFsara:

If you have any questions concerning the technical specifications below, please contact us at helpdesk@surfsara.nl.

Quick summary

Gina cluster Capacity
Operating system Linux CentOS 7.x 64bit
Total number of cores 7408 Xeon cores at 2.2 to 2.6 GHz
Total memory 58.3TB
Total scratch space 2258TB

Last update: June 2018

Network

GinA and Grid storage share a network backbone, connected with a high throughput network. Currently we have seen over 20GB/sec (~170-200Gbit/sec) of peak network traffic to the Grid storage. All workernodes are connected with a single 10Gbit or 25Gbit ethernet connection.

Worker Nodes

This is the list of the different worker nodes in order of installation/configuration from newest to oldest:

Worker Nodes f18-{01-12}, f19-{01-11}, f20-{01-12}:

*  Dell R640
*  2x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (40 cores)
*  384GB RAM
*  3TB scratch
*  Type: Skylake architecture
*  Number of nodes: 35
*  Total cores of this type: 1400
*  Scratch per core: ~80GB
*  RAM per core: 8GB

Worker Nodes f08-{01-16}, f09-{01-16}, f10-{01-16}, f11-{01-16}, f12-{01-17}:

*  Fujitsu RX2530
*  2x Intel(R) Xeon(R) CPU E5-2650 v4 @ 2.20GHz (24 cores)
*  192GB RAM
*  ~2.4TB scratch (SSD only)
*   Type: Broadwell architecture /w Broadwell-EP cores (2016)
*  Number of nodes: 81
*  Total cores of this type: 1944
*  Scratch per core: ~100GB
*  RAM per core: 8GB

Worker Nodes f03-{01-20}, f04-{01-20}, f05-{01-20}, f06-{01-20}, f07-{01-20}:

*  Dell R630
*  2x Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz (24 cores)
*  192GB RAM
*  ~8000GB scratch
*  Type: Haswell architecture /w Haswell-EP cores (2014)
*  Number of nodes: 100
*  Total cores of this type: 2400
*  Scratch per core: ~300GB
*  RAM per core: 8GB

Worker Nodes f13-{01-24}, f14-{01-20}, f15-{01-20}, f16-{01-20}, f17-{01-20}:

*  Fujitsu CX250
*  2x Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz (16 cores)
*  128GB RAM
*  ~11TB scratch
*  Type: Ivy Bridge architecture /w Ivy Bridge-EP cores (2013)
*  Number of nodes: 104
*  Total cores of this type: 1664
*  Scratch per core: ~680GB
*  RAM per core: 8GB

Service nodes

Service{01,02,03}:

*  Dell R420
*  2xIntel(R) Xeon(R) CPU E5-2420 0 @ 1.90GHz (12 cores)
*  96GB RAM
*  Number of nodes: 3
*  RAM per core: 8GB

CreamCEs

All 3 CreamCEs are virtualized and distributed among the 3 Service Nodes. Every CreamCE has 4 cores and 9GB RAM in total.

Queues

Queue Max. Walltime (hh:mm:ss) VOs(group) allowed
long 96:00:00 geant4 atlas(production) projects.nl(tropomi) virgo lsgrid(mine)
medium 36:00:00 astron atlas alice auger bbmri.nl beapps biomed dans enmr.eu esr geant4 lhcb lofar lsgrid nlesc.nl omegac pvier xenon.biggrid.nl projects.nl chem.biggrid.nl euclid-ec.org drihm.eu virgo skatelescope.eu dune
short 04:00:00 astron atlas auger bbmri.nl beapps biomed dans enmr.eu esr geant4 lhcb lofar lsgrid nlesc.nl omegac pvier xenon.biggrid.nl projects.nl chem.biggrid.nl euclid-ec.org drihm.eu dune
infra 00:30:00 dteam ops pvier

Last update: February 2019

The Grid jobs submitted to the queues above are restricted by the walltime limit, not the CPU limit.

Information on the queues available on the cluster at Nikhef can be found here