may 24-26 2004 s. timm--infrastructure and provisioning at fermilab hdcf 1 infrastructure and...
TRANSCRIPT
May 24-26 2004 S. Timm--Infrastructure and Provisioning at Fermilab HDCF
1
Infrastructure and Provisioning at the Fermilab High Density
Computing Facility
Steven C. Timm
Fermilab
HEPiX conference
May 24-26, 2004
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
2
Outline
Current Fermilab facilities Expected need for future Fermilab facilities Construction activity at High Density
Computing Facility Networking and power infrastructure Provisioning and management at remote
location
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
3
A cast of thousands….
HDCF design done by Fermilab Facilities Engineering, Construction by outside contractor Managed by CD Operations (G. Bellendir et al) Requirements planning by taskforce of Computing
Division personnel including system administrators, department heads, networking, facilities people.
Rocks development work by S. Timm, M. Greaney, J. Kaiser
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
4
Current Fermilab facilities
Feynman Computing Center built in 1988 (to house large IBM-compatible mainframe).
~18000 square feet of computer rooms ~200 tons of cooling Maximum input current 1800A Computer rooms backed up with UPS Full building backed up with generator ~1850 dual-CPU compute servers, ~200 multi-TB IDE RAID
servers in FCC right now Many other general-purpose servers, file servers, tape robots.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
5
Current facilities continued
Satellite computing facility in former experimental hall “New Muon Lab”
Historically for Lattice QCD clusters (208512) Now contains >320 other nodes waiting for
construction of new facility
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
6
The long hot summer
In summer it takes considerably more energy to run the air conditioning.
Dependent on shallow pond for cooling water. In May building has already come within 25A (out of
1800) from having to shut down equipment to shed power load and avoid brownout.
Current equipment exhausts the cooling capacity of Feynman computing center as well as the electric
No way to increase either in existing building without long service outages.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
7
Computers just keep getting hotter
Anticipate that in fall ’04 we can buy dual Intel 3.6 GHz “Nocona” chip, ~105W apiece
Expect at least 2.5A current draw per node, maybe more, 12-13 kVA per rack of 40 nodes.
In FCC we have 32 computers per rack, 8-9 kVA Have problems cooling the top nodes even now. New facility will have 5x more cooling, 270 tons for
2000 square feet New facility will have up to 3000A of electrical current
available.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
8
We keep needing more computers Moore’s law doubling time isn’t holding true in commodity market Computing needs are growing faster than Moore’s law and must be met with more computers 5 year projections are based on plans from experiments.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
9
Fermi Cycles as a function of time
Fermi cycles vs time
0
500
1000
1500
2000
2500
3000J-
98
D-9
8
J-9
9
D-9
9
J-0
0
D-0
0
J-0
1
D-0
1
J-0
2
D-0
2
J-0
3
D-0
3
Time
Fer
mi C
ycle
s
Fermi Cycles
Moore's law F=2.02
Y=R*2^(X/F): Moore’s law says F=1.5 years,
F=2.02 years and growing. 1000 Fermi Cycles PIII 1 GHz
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
10
Fermi Cycles per ampere as function of time
Performance per Ampere
0
500
1000
1500
2000
2500
9/1
/00
1/1
/01
5/1
/01
9/1
/01
1/1
/02
5/1
/02
9/1
/02
1/1
/03
5/1
/03
9/1
/03
1/1
/04
Time
Perfo
rm
an
ce/a
mp
ere
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
11
Fermi cycles per dollar as function of time
Performance/price
0.00
0.50
1.00
1.50
2.00
2.50
3.00
J-9
8
J-9
9
J-0
0
J-0
1
J-0
2
J-0
3
Time
Fer
mi c
ycle
s / $
US
Performance/price
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
12
Strategy:
Feynman center will be UPS and generator-backed facility for important servers
New HDCF will have UPS for graceful shutdown but no generator backup. Designed for high-density compute nodes (plus a few tape robots).
10-20 racks of existing 1U will be moved to new facility and reracked
Anticipate10-15 racks of new purchase this fall also in new building
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
13
Location ofHDCF
1.5 miles away from FCC
No administrators will be housed there—will manage “lights out”
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
14
Floor plan of HDCF
Room for 72 racks in each of 2 computer rooms.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
15
Cabling plan
Network Infrastructure
Will use bundles of individual Cat-6 cables
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
16
Current status
Construction began early May
Occupancy Nov/Dec 2004 (est).
Phase I+II, space for 56 racks at that time.
Expected cost: US$2.8M.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
17
Power/console infrastructure
Cyclades AlterPath series Includes console servers, network-based KVM
adapters, and power strips Alterpath ACS48 runs PPC Linux Supports Kerberos 5 authentication Access control can be divided by each port Any number of power strip outlets can be associated
with each machine on each console port. All configurable via command line or Java-based GUI
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
18
Power/console infrastructure
PM-10 Power strip
120VAC 30A
10 nodes/circuit
Four units/rack
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
19
Installation with NPACI-Rocks
NPACI (National Partnership for Advanced Computational Infrastructure), lead institution is San Diego Supercomputing Center
Rocks—”ultimate cluster-in-a-box tool.” Combines Linux distribution, database, highly modified installer, and a large amount of parallel computing applications such as PBS, Maui, SGE, MPICH, Atlas, PVFS.
Rocks 3.0 based on Red Hat Linux 7.3 Rocks 3.1 and greater based on SRPMS of Red Hat
Enterprise Linux 3.0.
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
20
Rocks vs. Fermi Linux comparison
REDHAT7.3
Adds:
Workgroups
Yum
OpenAFS
Fermi Kerberos/
OpenSSH
Adds:
Extended kickstart
HPC applications
MySQL database
Fermi Linux REDHAT Rocks 3.0
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
21
Rocks Fermiarchitecture Application
Expects all compute nodes on private net behind a firewall
Reinstall node if any changes
All network services (DHCP, DNS, NIS) supplied by the frontend.
Nodes on public net Users won’t allow
downtime for frequent reinstall
Use yum and other Fermi Linux tools for security updates
Configure ROCKS to use our external network services
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
22
Fermi extensions to Rocks
Fermi production farms currently have 752 nodes all installed with Rocks
This Rocks cluster has the most CPU’s registered of any cluster at rocksclusters.org
Added extra tables to database for customizing kickstart configuration (we have 14 different disk configurations)
Added Fermi Linux comps files to have all Fermi workgroups available in installs, and all added Fermi RPMS
Made slave frontend install servers during mass reinstall phases. During normal operation one install server is enough.
Added logic to recreate kerberos keytabs
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
23
S.M.A.R.T Monitoring
“smartd” daemon from smartmontools package gives early warning of disk failures
Disk failures are ~70% of all hardware failures in our farms over last 5 years.
Run short self-test on all disks every day
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
24
Temperature/power monitoring
Wrappers for lm_sensors feed NGOP and Ganglia. Measure average temperature of nodes over a month Alarm when 5C or 10C above average Page when 50% of any group at 10C above average Automated shutdown script activates when any single
node is over emergency temperature. Building-wide signal will provide notice that we are on
UPS power and have 5 minutes to shut down. Automated OS shutdown and SNMP poweroff scripts
May 24-26 2004S. Timm--Infrastructure and Provisioning at Fermilab HDCF
25
Reliability is key
Can only successfully manage remote clusters if hardware is reliable
All new contracts are written with vendor providing 3 year warranty parts and labor…they only make money if they build good hardware
30-day acceptance test is critical to identify hardware problems and fix them before production begins.
With 750 nodes and 99% reliability, still 8 nodes would be down a day.
Historically reliability is closer to 96% but new Intel-based Xeon nodes are much better.