- pacman -S vim dialog net-tools sudo zsh
- pacman -S xorg-server xorg-xinit xorg-utils xorg-server-utils xf86-video-intel xf86-input-evdev xf86-input-synaptics
- useradd -m -s /usr/bin/zsh cmaureir
Hosts | Batch Role | Grid Role |
---|---|---|
ui01..ui02 | PBS client | User Interface (UI) |
site | PBS server | Site BDII |
wn01..wn24 | PBS worker | Worker Nodes (WN) |
gp01..gp09 | PBS worker GPU | - |
ce01..ce02 | - | CREAM CE |
fs01..fs06 | File Servers | Storage Pools (Gluster) |
sp01..sp03 | - | Storage Element (dCache SE Pools) |
se | - | Storage Element (dCache SE Head) |
Calculate dt_{b_{i}}
according to Eq. 22
Calculate the \eta_b
using the condition: (Page 8, right column, second paragaph)
ETA_B = D_TIME_MIN / (2.0 * min(dt_{b_{i}});
Initialization of dt = D_TIME_MIN / 2 (Question1: This for everyone? or we just use the minimum?)
// ... | |
if(P[j].Mass > 0) | |
{ | |
if(mass > 0) | |
{ | |
dx = pos[0] - P[j].Pos[0]; | |
dy = pos[1] - P[j].Pos[1]; | |
dz = pos[2] - P[j].Pos[2]; | |
r2 = dx * dx + dy * dy + dz * dz; |
Category | State | SGE Letter Code |
---|---|---|
Pending | pending | qw |
Pending | pending, user hold | qw |
Pending | pending, system hold | hqw |
Pending | pending, user and system hold | hqw |
Pending | pending, user hold, re-queue | hRwq |
Pending | pending, system hold, re-queue | hRwq |
Source: http://supercomputers.aei.mpg.de/user-guide-2/lustre-file-system-acls-quota/
How to give permissions to a certain directory inside your home:
710
chmod 710 /home/youruser
externaluser
to a directory
inside your home (This can be rwx
, rw
, r-x
, etc.)Vamos a considerar las GPUs nuevas (Tesla K20m)
, por lo cual es bueno entender un par de detalles acerca del Hardware:
La RAM de la GPU (Memoria Global) es 4800 MB
, a esta memoria acceden todas las hebras de las funciones que llamemos,
pero el acceso es lento, pues es caro a nivel de hardware, por lo mismolas hebras tienen unas memorias chicas a las cuales
pueden acceder más rápido, las cuales no están dentro de la memoria global; éstas son los Registros y la Memoria compartida.
notify_user = <email> | |
notification = Error | |
notification = Complete | |
notification = Always | |
universe = vanilla | |
environment = OUTDIR=.;LD_LIBRARY_PATH=\${LD_LIBRARY_PATH}:<extra_path> | |
executable = <route to executable> | |
arguments = <arguments> | |
output = <output_file> |
#!/bin/bash | |
$ -r n | |
$ -l h_rt=24:00:00, h=!(<blacklist_node1>| ...) | |
$ -q <queue>.q | |
$ -pe <queue> 60 | |
$ -N <name> | |
$ -m abe | |
$ -cwd | |
$ -R y | |
* 2 GPUs (Tesla C1060) | |
* 24 CPUs (Intel(R) Xeon(R) X5650 @ 2.67GHz) | |
CUDA Driver Version / Runtime Version 6.0 / 5.5 | |
CUDA Capability Major/Minor version number: 2.0 | |
Total amount of global memory: 2687 MBytes (2817982464 bytes) | |
(14) Multiprocessors, ( 32) CUDA Cores/MP: 448 CUDA Cores | |
GPU Clock rate: 1147 MHz (1.15 GHz) | |
Memory Clock rate: 1546 Mhz | |
Memory Bus Width: 384-bit |