Cluster 1
Our cluster at the University of Toronto St George Campus in the Faculty of Pharmacy is described on this page. We also have Cluster 0 and Cluster 2 at UCSF in San Francisco.
Status
Cluster 1 is being wound down. Please plan to move your files to Cluster 2 in May 2014.
Toronto - our cluster
Obtaining credentials
For command line access to our cluster, please contact Teague, Therese or John for credentials. For your main email address, you may well use your gmail account or other public account for most purposes.
Disk usage conventions
We differentiate dockers from everyone else. If you are a full time member of the group, your disk space allocation will be like this:
Your home directory is for important or small files. 500 GB default. Backed up.
/nfs/home/<userid>/
We expect you to run docking calculations in work. 1TB default. Not backed up.
/nfs/work/<userid>/
If you have large amounts of data you want kept, e.g. from the synchrotron, you may request repository space to keep it. Backed up. Size by agreement. Please compress your image files!
/nfs/store/<userid>/
If you generate private databases, you may request /db/ space. Backup by arrangement. Size by agreement.
/nfs/db/<userid>/
If you have software you want to share, you may request /software/ space. backed up. 500 GB.
/nfs/software/<userid>/
Getting set up with lab software
svn co dockenv cp $SOFT/labenv.csh ~/.csh
OR
cp $SOFT/labenv.sh ~/.bashrc
for bash users.
Roles and Functions of Computers
routine access to cluster
- sgehead. includes csd license. pgf license
pipeline pilot license
- ppilot
Sysadmin
static web server
- runaway (vm on rocky)
- includes wiki, www
- NB was contact at some point, kinda liked that name for a webserver
- includes drupal
- includes wordpress
postgres server
- production: taxi
- intern: raiders
dynamic web server (ZINC, etc)
- oz
SEA server
- fawlty
- mysql server is on msqlserver aka inception
- fingerprint server is on fingerprint aka darkcrystal
NFS Disk servers
- fawlty
- towers
Name reuse pool
- vertigo
- sundance
- spotless
- lostark
- airplane
- brian
- terminator
- psycho
- strangelove
Cloud machines (hypervisor)
- Django
- Pulp
- Raiders
- Strangelove
- Rocky
Unknown/unassigned function
- blues
- goldmember
- darkness
389 Server
- spinaltap
Physical Machines
NB this section needs updating!
Rack 1
Currently in Donnelly 630. Totals: 426 cores for cluster. 128 cores for infrastructure. 60 TB RAID-10 formatted. (130 TB raw)
- 3x ( Dell PowerEdge C6145. = 2 x ( 256 GB / 64 cores / 6 TB raw disk )) 64 x 2 x 3 = 384 cores.
- HP disk head node, (25 TB raw disk systems ) x 4.
- Dell disk head node x 1, Dell SAS 2TB x 8??.
- DL165 x 4. 8x4 = 32 cores.
- Dell Switch: switch-1-1.
- UPS
Rack 2
Currently in Pharmacy 8xx. Planned June 2013
- Dell PowerEdge C6145
- Dell Switch: switch-2-1.
- 42 tape robot backup system
- dedicated computer for backup
Rack 3
Currently in Pharmacy 8xx. Planned June 2013
- Dell PowerEdge C6145
- Dell Switch: switch-3-1.
- legacy nodes
Rack 4
Currently in Pharmacy 8xx. Planned June 2013
- will be: Dell PowerEdge 6145
- Dell Switch: switch-4-1.
- legacy nodes
Rack 5
Currently in Pharmacy 8xx. Planned June 2013
- will be: Dell PowerEdge 6145
- Dell Switch: switch-5-1.
- legacy nodes
Wet Lab
Donnelly 940. Planned June 2013.
- 4 instrument workstations
- 3 linux desktops HP Z430
- 1 windows7 desktop HP Z400
- 1 color printer
- 3 Mac lab laptops
- 1 switch: switch-w-1
Dry Lab
Donnelly 650. Planned June 2013.
- 6 linux desktops. HP Z430
- 1 windows7 desktop HP Z400
- 6 Mac lab laptops
- 1 switch: switch-d-1
Other Offices and locations
- BKS office
- JJI office
- desk space in OICR