https://wiki.cs.earlham.edu/api.php?action=feedcontributions&user=Skylar&feedformat=atom Earlham CS Department - User contributions [en] 2024-03-29T06:36:25Z User contributions MediaWiki 1.32.1 https://wiki.cs.earlham.edu/index.php?title=Al-salam&diff=10648 Al-salam 2010-01-28T01:21:55Z <p>Skylar: /* Quick breakdown */</p> <hr /> <div>Al-Salam is the working name for the Earlham Computer Science Department's upcoming cluster computer.<br /> <br /> At the moment Al-Salam exists only as a $40,000 grant and a growing list of tentative specifications:<br /> <br /> == Latest Overarching Questions ==<br /> *Should we build this machine ourselves?<br /> *#Are we wasting our money and learning opportunity letting them do the building for us?<br /> *#If it is cheaper, Would it be a useful experience for the students this coming semester to take a large collection of hardware and make it into a cluster?<br /> *# Yes, this was pretty clear from the email thread in early December.<br /> *How much if any GPGPU hardware do we want? 0, 1 or 2 nodes worth?<br /> *Do we want a high bandwidth/low latency network?<br /> **We do not. More expensive than it is worth.<br /> *What software stack do we want to run? Vendor supplied or the BCCD?<br /> **Both. Vendor-supplied base with a BCCD virtual machine<br /> ***Will the virtual machine support CUDA?<br /> * Do compute nodes have spinning disk? <br /> ** Compute nodes have a spinning disk. Solid state is still expensive<br /> * What's on the local persistent store? /tmp? An entire OS?<br /> * Support<br /> ** Consider getting the cheapest hardware support, loosing a node isn't critical as long as they send a replacement quickly.<br /> <br /> == Parts List ==<br /> # Nodes - case, motherboard(s), power supply, CPU, RAM, GPGPU cards<br /> # Switch - managed, cut-through<br /> # Power distribution - rack-mount PDUs<br /> <br /> == Tentative Specifications ==<br /> <br /> === Budget ===<br /> * $35,000 (leaving $5,000 for discretionary spending)<br /> <br /> === Nodes ===<br /> * Intel Nehalem processors<br /> * 4 core processors minimum<br /> ** Six cores still expensive<br /> * 1.5GB RAM per core<br /> <br /> === Specialty Nodes ===<br /> * Two nodes should support CUDA GPGPU<br /> <br /> Educationally, we could expect to get significant use out of GPGPUs, but the production use is limited.<br /> Increasing the variance of the architecture landscape would be a bonus to education.<br /> <br /> === Network ===<br /> * Gigabit Ethernet fabric with switch<br /> <br /> === Disk ===<br /> * Spinning Disk<br /> <br /> === OS ===<br /> * Virtual BCCD on top of built-in OS.<br /> <br /> == Quick breakdown ==<br /> <br /> === Nodes ===<br /> <br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> !<br /> ! [[Al-salam#ION_Computer_Systems_Quotation_.2361116|ION #61116]]<br /> ! [[Al-salam#ION_Computer_Systems_Quotation_.2361164|ION #61164]]<br /> ! [[Al-salam#Silicon_Mechanics_Quote_.23174536|SM #174536]]<br /> ! [[Al-salam#Newegg_Quote_.231|Newegg #1]]<br /> ! [[Al-salam#Newegg_Quote_.232|Newegg #2]]<br /> |-<br /> | '''CPU'''<br /> | 72 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> | 128 2.4GHz Intel E5530<br /> | 112 2.4GHz Intel E5530<br /> |-<br /> | '''RAM'''<br /> | 108GB PC3-10600<br /> | 120GB PC3-10600<br /> | 120GB DDR3-1333<br /> | 192GB DDR3-1333<br /> | 168GB DDR3-1333<br /> |-<br /> | '''GPU'''<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> | None<br /> | 4 Tesla C1060<br /> |-<br /> | '''Local disk'''<br /> | Yes<br /> | Yes<br /> | Yes<br /> | Yes<br /> | Yes<br /> |-<br /> | '''Shared chassis'''<br /> | No<br /> | Yes<br /> | Yes<br /> | No<br /> | No<br /> |-<br /> | '''Remote mgmt'''<br /> | No<br /> | No<br /> | IPMI<br /> | No<br /> | IPMI on GPU nodes<br /> |-<br /> | '''Size (just nodes)'''<br /> | 9U<br /> | 6U<br /> | 6U<br /> | 16U<br /> | 12U<br /> |-<br /> | '''Price'''<br /> | $33,173.20<br /> | $33,054.30<br /> | $30,078.00<br /> | $32,910.56<br /> | $34,696.78<br /> |}<br /> <br /> === Power distribution ===<br /> <br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> !<br /> ! [http://www.tripplite.com/en/products/model.cfm?txtSeriesID=446&amp;EID=14295&amp;txtModelID=2005 PDU1220]<br /> ! [http://www.tripplite.com/en/products/model.cfm?txtSeriesID=513&amp;EID=77300&amp;txtModelID=3867 PDUMH20]<br /> ! [http://accessories.us.dell.com/sna/products/Server_Network/productdetail.aspx?c=us&amp;l=en&amp;cs=04&amp;sku=A0151958 AP9563]<br /> ! [http://accessories.us.dell.com/sna/products/Server_Network/productdetail.aspx?c=us&amp;l=en&amp;s=bsd&amp;cs=04&amp;sku=A0748689 AP7801]<br /> |-<br /> | '''Vendor'''<br /> | TrippLite<br /> | TrippLite<br /> | APC<br /> | APC<br /> |-<br /> | '''Size'''<br /> | 1U<br /> | 1U<br /> | 1U<br /> | 1U<br /> |-<br /> | '''Capabilities'''<br /> | Dumb<br /> | Metered<br /> | Dumb<br /> | Metered<br /> |-<br /> | '''Input power'''<br /> | 20A, 1x NEMA 5-20P<br /> | 20A, 1x NEMA L5-20P w/ NEMA 5-20P adapter<br /> | 20A, 1x NEMA 5-20P<br /> | 20A, 1x NEMA 5-20P<br /> |-<br /> | '''Output power'''<br /> | 13x NEMA 5-20R<br /> | 12x NEMA 5-20R<br /> | 10x NEMA 5-20R<br /> | 8x NEMA 5-20R<br /> |-<br /> | '''Price'''<br /> | $195<br /> | $230<br /> | $120<br /> | $380<br /> |}<br /> <br /> == ION Computer Systems Quotation #61116 ==<br /> *2 ION G10 Server with GPU: 4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> *7 ION G10 Server without GPU: $3,697.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * Networking Fabric<br /> **Network not included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> * Price tag: $33,173.20<br /> <br /> == ION Computer Systems Quotation #61164 ==<br /> <br /> * 2 ION G10 Server with GPU: $4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * 4 ION T11 DualNode: $6,477.0 each<br /> ** (2x2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** Total memory: 12GB DDR3_1333 per node<br /> ** No RAID, Separate disks (NO redundancy)<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate Constellation 160GB, 7200RPM, SATA 3Gb NCQ 2.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> ** These nodes are modular. One can be unplugged and worked on while the others remain running.<br /> <br /> * Network<br /> ** Network Not Included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> *Price Tag: $33,054.30<br /> <br /> ==Silicon Mechanics Quote #174536==<br /> * 2x Rackform iServ R4410: $11043.00 each ($10601.00 each with education) [http://www.siliconmechanics.com/quotes/174536?confirmation=879366549 link]<br /> **Shared Chassis: The following chassis resources are shared by all 4 compute nodes<br /> **External Optical Drive: No Item Selected<br /> **Power Supply: Shared, Redundant 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: Quick-Release Rail Kit for Square Holes, 26.5 - 36.4 inches<br /> **Compute Nodes x4<br /> ***CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> ***RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> ***NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> ***Management: Integrated IPMI with KVM over LAN<br /> ***Hot-Swap Drive - 1: 250GB Western Digital RE3 (3.0Gb/s, 7.2Krpm, 16MB Cache) SATA<br /> <br /> * 2x Rackform iServ R350-GPU: $5196.00 each ($4433.00 each with education) [http://www.siliconmechanics.com/quotes/174542?confirmation=712641984 link]<br /> **CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> **RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> **NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> **Management: Integrated IPMI 2.0 &amp; KVM with Dedicated LAN<br /> **GPU: 1U System with 1 x Tesla C1060 GPU, Actively Cooled<br /> **LP PCIe x4 2.0 (x16 Slot): No Item Selected<br /> **Hot-Swap Drive - 1: 250GB Seagate Barracuda ES.2 (3Gb/s, 7.2Krpm, 32MB Cache, NCQ) SATA<br /> **Power Supply: 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: 1U Rail Kit<br /> <br /> *Price Tag: $32,478 ($30,078 with education)<br /> <br /> *Questions<br /> **Can we lose the hot-swappability to save money?<br /> **Do we need to get a Gig-Switch?<br /> ***Would Cairo do?<br /> <br /> ==Newegg Quote #1==<br /> * 16x [http://secure.newegg.com/WishList/PublicWishDetail.aspx?WishListNumber=16958187 Newegg list]<br /> ** 1U [http://www.newegg.com/Product/Product.aspx?item=N82E16811116011 link]<br /> ** 2x Intel Xeon (Nehalem) E5530, Quad-Core, 2.4GHz, 80 Watt [http://www.newegg.com/Product/Product.aspx?item=N82E16819117184 link]<br /> ** Slim CD/DVD Drive<br /> ** 4x Gigabit ethernet [http://www.newegg.com/Product/Product.aspx?item=N82E16813151195R motherboard]<br /> ** 500W non-redundant power supply<br /> ** 160GB 7200RPM Seagate [http://www.newegg.com/Product/Product.aspx?item=N82E16822148511 link]<br /> ** 12GB RAM (240-pin DDR3 1333 ECC, unbuffered)<br /> <br /> * Price Tag: $32,910.56<br /> <br /> ==Newegg Quote #2==<br /> * 2x [http://secure.newegg.com/WishList/PublicWishDetail.aspx?WishListNumber=8560589 Newegg list]<br /> ** 1U<br /> ** 2x Intel Xeon (Nehalem) E5530, Quad-Core, 2.4GHz, 80 Watt<br /> ** 2x Gigabit ethernet<br /> ** IPMI<br /> ** 1400W non-redundant power supply<br /> ** 2x C1060 Tesla<br /> ** 160GB 7200RPM Seagate<br /> ** 12GB RAM (240-pin DDR3 1333 ECC, unbuffered)<br /> <br /> * 12x [http://secure.newegg.com/WishList/PublicWishDetail.aspx?WishListNumber=16958187 Newegg list]<br /> ** 1U [http://www.newegg.com/Product/Product.aspx?item=N82E16811116011 link]<br /> ** 2x Intel Xeon (Nehalem) E5530, Quad-Core, 2.4GHz, 80 Watt<br /> ** Slim CD/DVD Drive<br /> ** 4x Gigabit ethernet<br /> ** 500W non-redundant power supply<br /> ** 160GB 7200RPM Seagate<br /> ** 12GB RAM (240-pin DDR3 1333 ECC, unbuffered)<br /> <br /> * Price tag: $34,696.78</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Al-salam&diff=10579 Al-salam 2009-12-17T05:15:12Z <p>Skylar: /* Quick breakdown */</p> <hr /> <div>Al-Salam is the working name for the Earlham Computer Science Department's upcoming cluster computer.<br /> <br /> At the moment Al-Salam exists only as a $40,000 grant and a growing list of tentative specifications:<br /> <br /> == Latest Overarching Questions ==<br /> *Should we build this machine ourselves?<br /> *#Are we wasting our money and learning opportunity letting them do the building for us?<br /> *#If it is cheaper, Would it be a useful experience for the students this coming semester to take a large collection of hardware and make it into a cluster?<br /> *How much if any GPGPU hardware do we want? 0, 1 or 2 nodes worth?<br /> *Do we want a high bandwidth/low latency network?<br /> **We do not. More expensive than it works<br /> *What software stack do we want to run? Vendor supplied or the BCCD?<br /> **Both. Vendor-supplied base with a BCCD virtual machine<br /> * Do compute nodes have spinning disk? <br /> ** Compute nodes have a spinning disk<br /> * What's on the local persistent store? /tmp? An entire OS?<br /> <br /> == Tentative Specifications ==<br /> <br /> === Budget ===<br /> * $35,000 (leaving $5,000 for discretionary spending)<br /> <br /> === Nodes ===<br /> * Intel Nehalem processors<br /> * 4 core processors minimum<br /> ** Six cores still expensive<br /> * 1.5GB RAM per core<br /> <br /> === Specialty Nodes ===<br /> * Two nodes should support CUDA GPGPU<br /> <br /> Educationally, we could expect to get significant use out of GPGPUs, but the production use is limited.<br /> Increasing the variance of the architecture landscape would be a bonus to education.<br /> <br /> === Network ===<br /> * Gigabit Ethernet fabric with switch<br /> <br /> === Disk ===<br /> * Spinning Disk<br /> <br /> === OS ===<br /> * Virtual BCCD on top of built-in OS.<br /> <br /> == Quick breakdown ==<br /> <br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> !<br /> ! [[Al-salam#ION_Computer_Systems_Quotation_.2361116|ION #61116]]<br /> ! [[Al-salam#ION_Computer_Systems_Quotation_.2361164|ION #61164]]<br /> ! [[Al-salam#Silicon_Mechanics_Quote_.23174536|SM #174536]]<br /> |-<br /> | '''CPU'''<br /> | 72 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> |-<br /> | '''RAM'''<br /> | 108GB PC3-10600<br /> | 120GB PC3-10600<br /> | 120GB DDR3-1333<br /> |-<br /> | '''GPU'''<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> |-<br /> | '''Local disk'''<br /> | Yes<br /> | Yes<br /> | Yes<br /> |-<br /> | '''Shared chassis'''<br /> | No<br /> | Yes<br /> | Yes<br /> |-<br /> | '''Remote mgmt'''<br /> | No<br /> | No<br /> | IPMI<br /> |-<br /> | '''Price'''<br /> | $33,173.20<br /> | $33,054.30<br /> | $30,078.00<br /> |}<br /> <br /> == ION Computer Systems Quotation #61116 ==<br /> *2 ION G10 Server with GPU: 4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> *7 ION G10 Server without GPU: $3,697.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * Networking Fabric<br /> **Network not included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> * Price tag: $33,173.20<br /> <br /> == ION Computer Systems Quotation #61164 ==<br /> <br /> * 2 ION G10 Server with GPU: $4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * 4 ION T11 DualNode: $6,477.0 each<br /> ** (2x2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** Total memory: 12GB DDR3_1333 per node<br /> ** No RAID, Separate disks (NO redundancy)<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate Constellation 160GB, 7200RPM, SATA 3Gb NCQ 2.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> ** These nodes are modular. One can be unplugged and worked on while the others remain running.<br /> <br /> * Network<br /> ** Network Not Included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> *Price Tag: $33,054.30<br /> <br /> ==Silicon Mechanics Quote #174536==<br /> * 2x Rackform iServ R4410: $11043.00 each ($10601.00 each with education) [http://www.siliconmechanics.com/quotes/174536?confirmation=879366549 link]<br /> **Shared Chassis: The following chassis resources are shared by all 4 compute nodes<br /> **External Optical Drive: No Item Selected<br /> **Power Supply: Shared, Redundant 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: Quick-Release Rail Kit for Square Holes, 26.5 - 36.4 inches<br /> **Compute Nodes x4<br /> ***CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> ***RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> ***NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> ***Management: Integrated IPMI with KVM over LAN<br /> ***Hot-Swap Drive - 1: 250GB Western Digital RE3 (3.0Gb/s, 7.2Krpm, 16MB Cache) SATA<br /> <br /> * 2x Rackform iServ R350-GPU: $5196.00 each ($4433.00 each with education) [http://www.siliconmechanics.com/quotes/174542?confirmation=712641984 link]<br /> **CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> **RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> **NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> **Management: Integrated IPMI 2.0 &amp; KVM with Dedicated LAN<br /> **GPU: 1U System with 1 x Tesla C1060 GPU, Actively Cooled<br /> **LP PCIe x4 2.0 (x16 Slot): No Item Selected<br /> **Hot-Swap Drive - 1: 250GB Seagate Barracuda ES.2 (3Gb/s, 7.2Krpm, 32MB Cache, NCQ) SATA<br /> **Power Supply: 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: 1U Rail Kit<br /> <br /> *Price Tag: $32,478 ($30,078 with education)<br /> <br /> *Questions<br /> **Can we lose the hot-swappability to save money?<br /> **Do we need to get a Gig-Switch?<br /> ***Would Cairo do?</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Al-salam&diff=10578 Al-salam 2009-12-17T05:13:58Z <p>Skylar: /* Table */</p> <hr /> <div>Al-Salam is the working name for the Earlham Computer Science Department's upcoming cluster computer.<br /> <br /> At the moment Al-Salam exists only as a $40,000 grant and a growing list of tentative specifications:<br /> <br /> == Latest Overarching Questions ==<br /> *Should we build this machine ourselves?<br /> *#Are we wasting our money and learning opportunity letting them do the building for us?<br /> *#If it is cheaper, Would it be a useful experience for the students this coming semester to take a large collection of hardware and make it into a cluster?<br /> *How much if any GPGPU hardware do we want? 0, 1 or 2 nodes worth?<br /> *Do we want a high bandwidth/low latency network?<br /> **We do not. More expensive than it works<br /> *What software stack do we want to run? Vendor supplied or the BCCD?<br /> **Both. Vendor-supplied base with a BCCD virtual machine<br /> * Do compute nodes have spinning disk? <br /> ** Compute nodes have a spinning disk<br /> * What's on the local persistent store? /tmp? An entire OS?<br /> <br /> == Tentative Specifications ==<br /> <br /> === Budget ===<br /> * $35,000 (leaving $5,000 for discretionary spending)<br /> <br /> === Nodes ===<br /> * Intel Nehalem processors<br /> * 4 core processors minimum<br /> ** Six cores still expensive<br /> * 1.5GB RAM per core<br /> <br /> === Specialty Nodes ===<br /> * Two nodes should support CUDA GPGPU<br /> <br /> Educationally, we could expect to get significant use out of GPGPUs, but the production use is limited.<br /> Increasing the variance of the architecture landscape would be a bonus to education.<br /> <br /> === Network ===<br /> * Gigabit Ethernet fabric with switch<br /> <br /> === Disk ===<br /> * Spinning Disk<br /> <br /> === OS ===<br /> * Virtual BCCD on top of built-in OS.<br /> <br /> == Quick breakdown ==<br /> <br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> !<br /> ! ION #61116<br /> ! ION #61164<br /> ! SM #174536<br /> |-<br /> | '''CPU'''<br /> | 72 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> |-<br /> | '''RAM'''<br /> | 108GB PC3-10600<br /> | 120GB PC3-10600<br /> | 120GB DDR3-1333<br /> |-<br /> | '''GPU'''<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> |-<br /> | '''Local disk'''<br /> | Yes<br /> | Yes<br /> | Yes<br /> |-<br /> | '''Shared chassis'''<br /> | No<br /> | Yes<br /> | Yes<br /> |-<br /> | '''Remote mgmt'''<br /> | No<br /> | No<br /> | IPMI<br /> |-<br /> | '''Price'''<br /> | $33,173.20<br /> | $33,054.30<br /> | $30,078.00<br /> |}<br /> <br /> == ION Computer Systems Quotation #61116 ==<br /> *2 ION G10 Server with GPU: 4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> *7 ION G10 Server without GPU: $3,697.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * Networking Fabric<br /> **Network not included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> * Price tag: $33,173.20<br /> <br /> == ION Computer Systems Quotation #61164 ==<br /> <br /> * 2 ION G10 Server with GPU: $4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * 4 ION T11 DualNode: $6,477.0 each<br /> ** (2x2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** Total memory: 12GB DDR3_1333 per node<br /> ** No RAID, Separate disks (NO redundancy)<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate Constellation 160GB, 7200RPM, SATA 3Gb NCQ 2.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> ** These nodes are modular. One can be unplugged and worked on while the others remain running.<br /> <br /> * Network<br /> ** Network Not Included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> *Price Tag: $33,054.30<br /> <br /> ==Silicon Mechanics Quote #174536==<br /> * 2x Rackform iServ R4410: $11043.00 each ($10601.00 each with education) [http://www.siliconmechanics.com/quotes/174536?confirmation=879366549 link]<br /> **Shared Chassis: The following chassis resources are shared by all 4 compute nodes<br /> **External Optical Drive: No Item Selected<br /> **Power Supply: Shared, Redundant 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: Quick-Release Rail Kit for Square Holes, 26.5 - 36.4 inches<br /> **Compute Nodes x4<br /> ***CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> ***RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> ***NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> ***Management: Integrated IPMI with KVM over LAN<br /> ***Hot-Swap Drive - 1: 250GB Western Digital RE3 (3.0Gb/s, 7.2Krpm, 16MB Cache) SATA<br /> <br /> * 2x Rackform iServ R350-GPU: $5196.00 each ($4433.00 each with education) [http://www.siliconmechanics.com/quotes/174542?confirmation=712641984 link]<br /> **CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> **RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> **NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> **Management: Integrated IPMI 2.0 &amp; KVM with Dedicated LAN<br /> **GPU: 1U System with 1 x Tesla C1060 GPU, Actively Cooled<br /> **LP PCIe x4 2.0 (x16 Slot): No Item Selected<br /> **Hot-Swap Drive - 1: 250GB Seagate Barracuda ES.2 (3Gb/s, 7.2Krpm, 32MB Cache, NCQ) SATA<br /> **Power Supply: 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: 1U Rail Kit<br /> <br /> *Price Tag: $32,478 ($30,078 with education)<br /> <br /> *Questions<br /> **Can we lose the hot-swappability to save money?<br /> **Do we need to get a Gig-Switch?<br /> ***Would Cairo do?</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Al-salam&diff=10577 Al-salam 2009-12-17T05:11:40Z <p>Skylar: </p> <hr /> <div>Al-Salam is the working name for the Earlham Computer Science Department's upcoming cluster computer.<br /> <br /> At the moment Al-Salam exists only as a $40,000 grant and a growing list of tentative specifications:<br /> <br /> == Latest Overarching Questions ==<br /> *Should we build this machine ourselves?<br /> *#Are we wasting our money and learning opportunity letting them do the building for us?<br /> *#If it is cheaper, Would it be a useful experience for the students this coming semester to take a large collection of hardware and make it into a cluster?<br /> *How much if any GPGPU hardware do we want? 0, 1 or 2 nodes worth?<br /> *Do we want a high bandwidth/low latency network?<br /> **We do not. More expensive than it works<br /> *What software stack do we want to run? Vendor supplied or the BCCD?<br /> **Both. Vendor-supplied base with a BCCD virtual machine<br /> * Do compute nodes have spinning disk? <br /> ** Compute nodes have a spinning disk<br /> * What's on the local persistent store? /tmp? An entire OS?<br /> <br /> == Tentative Specifications ==<br /> <br /> === Budget ===<br /> * $35,000 (leaving $5,000 for discretionary spending)<br /> <br /> === Nodes ===<br /> * Intel Nehalem processors<br /> * 4 core processors minimum<br /> ** Six cores still expensive<br /> * 1.5GB RAM per core<br /> <br /> === Specialty Nodes ===<br /> * Two nodes should support CUDA GPGPU<br /> <br /> Educationally, we could expect to get significant use out of GPGPUs, but the production use is limited.<br /> Increasing the variance of the architecture landscape would be a bonus to education.<br /> <br /> === Network ===<br /> * Gigabit Ethernet fabric with switch<br /> <br /> === Disk ===<br /> * Spinning Disk<br /> <br /> === OS ===<br /> * Virtual BCCD on top of built-in OS.<br /> <br /> == Table ==<br /> <br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> !<br /> ! ION #61116<br /> ! ION #61164<br /> ! SM #174536<br /> |-<br /> | '''CPU'''<br /> | 72 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> | 80 2.4GHz Intel E5530<br /> |-<br /> | '''RAM'''<br /> | 108GB PC3-10600<br /> | 120GB PC3-10600<br /> | 120GB DDR3-1333<br /> |-<br /> | '''GPU'''<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> | 2 Tesla C1060<br /> |-<br /> |}<br /> <br /> == ION Computer Systems Quotation #61116 ==<br /> *2 ION G10 Server with GPU: 4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> *7 ION G10 Server without GPU: $3,697.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * Networking Fabric<br /> **Network not included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> * Price tag: $33,173.20<br /> <br /> == ION Computer Systems Quotation #61164 ==<br /> <br /> * 2 ION G10 Server with GPU: $4,972.00 each<br /> ** (2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** 12GB RAM [Bank 1 of 2: (6) 2GB ECC PC3-10600 1333MHz 2rank DDR3 RDIMM Modules][Smart]<br /> ** Total memory: 12GB DDR3_1333<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate SV35.3 250GB, 7200RPM, SATA 3Gb for SDVR 3.5“ Disk<br /> ** (1) NVidia Tesla C1060 w. 4GB DDR3<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> <br /> * 4 ION T11 DualNode: $6,477.0 each<br /> ** (2x2) Intel® Quad-Core Xeon® processor E5530 (2.40GHz, 8MB Cache, 5.86GT/s, 80W)<br /> ** Total memory: 12GB DDR3_1333 per node<br /> ** No RAID, Separate disks (NO redundancy)<br /> ** Configure 1 RAID sets / arrays.<br /> ** Seagate Constellation 160GB, 7200RPM, SATA 3Gb NCQ 2.5“ Disk<br /> ** Dual Intel Gigabit Server NICs with IOAT2 Integrated<br /> ** These nodes are modular. One can be unplugged and worked on while the others remain running.<br /> <br /> * Network<br /> ** Network Not Included<br /> <br /> * Other stuff<br /> ** scorpion: ION bootable USB Flash device for trouble shooting.<br /> ** 3 year Next Business Day response Onsite Repair Service by Source Support<br /> ** Default load for testing (Service Partition + CentOS 5.3 for Intel64)<br /> <br /> *Price Tag: $33,054.30<br /> <br /> ==Silicon Mechanics Quote #174536==<br /> * 2x Rackform iServ R4410: $11043.00 each ($10601.00 each with education) [http://www.siliconmechanics.com/quotes/174536?confirmation=879366549 link]<br /> **Shared Chassis: The following chassis resources are shared by all 4 compute nodes<br /> **External Optical Drive: No Item Selected<br /> **Power Supply: Shared, Redundant 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: Quick-Release Rail Kit for Square Holes, 26.5 - 36.4 inches<br /> **Compute Nodes x4<br /> ***CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> ***RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> ***NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> ***Management: Integrated IPMI with KVM over LAN<br /> ***Hot-Swap Drive - 1: 250GB Western Digital RE3 (3.0Gb/s, 7.2Krpm, 16MB Cache) SATA<br /> <br /> * 2x Rackform iServ R350-GPU: $5196.00 each ($4433.00 each with education) [http://www.siliconmechanics.com/quotes/174542?confirmation=712641984 link]<br /> **CPU: 2 x Intel Xeon E5530 Quad-Core 2.40GHz, 8MB Cache, 5.86GT/s QPI<br /> **RAM: 12GB (6 x 2GB) Operating at 1333MHz Max (DDR3-1333 ECC Unbuffered DIMMs)<br /> **NIC: Intel 82576 Dual-Port Gigabit Ethernet Controller - Integrated<br /> **Management: Integrated IPMI 2.0 &amp; KVM with Dedicated LAN<br /> **GPU: 1U System with 1 x Tesla C1060 GPU, Actively Cooled<br /> **LP PCIe x4 2.0 (x16 Slot): No Item Selected<br /> **Hot-Swap Drive - 1: 250GB Seagate Barracuda ES.2 (3Gb/s, 7.2Krpm, 32MB Cache, NCQ) SATA<br /> **Power Supply: 1400W Power Supply with PMBus - 80 PLUS Gold Certified<br /> **Rail Kit: 1U Rail Kit<br /> <br /> *Price Tag: $32,478 ($30,078 with education)<br /> <br /> *Questions<br /> **Can we lose the hot-swappability to save money?<br /> **Do we need to get a Gig-Switch?<br /> ***Would Cairo do?</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Talk:Al-salam&diff=10576 Talk:Al-salam 2009-12-17T04:59:05Z <p>Skylar: New page: * Re BCCD in a VM: Can CUDA be exported to a VM? If not, is that a disadvantage?</p> <hr /> <div>* Re BCCD in a VM: Can CUDA be exported to a VM? If not, is that a disadvantage?</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10201 Cluster:Modules 2009-09-04T03:40:55Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCHPATH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCHPATH --enable-shared &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCHPATH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCHPATH/lib --with-static=yes &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.<br /> ** After installation, changed the version from 3.2.6-&gt;3.2.7 in the init directory, and /usr/share/Modules to /cluster/software/modules-sw/modules/3.2.7/$ARCHPATH/Modules<br /> * OpenMPI: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/openmpi/1.3.1/$ARCHPATH &amp;&amp; make&lt;/code&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10200 Cluster:Modules 2009-09-04T03:30:51Z <p>Skylar: </p> <hr /> <div>= Software build options =<br /> <br /> ARCHPATH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCHPATH --enable-shared &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCHPATH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCHPATH/lib --with-static=yes &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.<br /> ** After installation, changed the version from 3.2.6-&gt;3.2.7 in the init directory, and /usr/share/Modules to /cluster/software/modules-sw/modules/3.2.7/$ARCHPATH/Modules</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10167 Cluster:Modules 2009-08-30T01:58:03Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCH --enable-shared &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib --with-static=yes &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.<br /> ** After installation, changed the version from 3.2.6-&gt;3.2.7 in the init directory, and /usr/share/Modules to /cluster/software/modules-sw/modules/3.2.7/$ARCH/Modules</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10166 Cluster:Modules 2009-08-30T01:21:47Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCH &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib --with-static=yes &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.<br /> ** After installation, changed the version from 3.2.6-&gt;3.2.7 in the init directory, and /usr/share/Modules to /cluster/software/modules-sw/modules/3.2.7/$ARCH/Modules</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10165 Cluster:Modules 2009-08-30T01:12:26Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCH &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.<br /> ** After installation, changed the version from 3.2.6-&gt;3.2.7 in the init directory, and /usr/share/Modules to /cluster/software/modules-sw/modules/3.2.7/$ARCH/Modules</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10164 Cluster:Modules 2009-08-30T01:02:02Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/tcl/8.5.7/$ARCH &amp;&amp; make&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/modules/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10163 Cluster:Modules 2009-08-30T00:59:22Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10162 Cluster:Modules 2009-08-30T00:57:28Z <p>Skylar: /* Software build options */</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;<br /> ** Had to remove bash_completion from init/Makefile.</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:Modules&diff=10161 Cluster:Modules 2009-08-30T00:54:06Z <p>Skylar: New page: = Software build options = ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p` * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules/modules-sw/3.2.7/$ARCH --with-tcl=/clu...</p> <hr /> <div>= Software build options =<br /> <br /> ARCH=`uname -s`/`/cluster/software/os_release`/`uname -p`<br /> <br /> * Tcl: &lt;code&gt;./configure --prefix=/cluster/software/modules/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;<br /> * Modules: &lt;code&gt;./configure --prefix=/cluster/software/modules/modules-sw/3.2.7/$ARCH --with-tcl=/cluster/software/modules-sw/tcl/8.5.7/$ARCH/lib &amp;&amp; make -j3&lt;/code&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster_Information&diff=10160 Cluster Information 2009-08-30T00:51:43Z <p>Skylar: /* General Stuff */</p> <hr /> <div>== Summer of Fun (2009) ==<br /> [[GalaxSee|An external doc for GalaxSee]]&lt;br /&gt;<br /> [[Cluster:OSGal|Documentation for OpenSim GalaxSee]]<br /> <br /> What's in the database?<br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> ! rowspan =&quot;2&quot; |<br /> ! colspan =&quot;3&quot; | GalaxSee (MPI) <br /> ! colspan =&quot;3&quot; | area-under-curve (MPI, openmpi) <br /> ! colspan =&quot;3&quot; | area-under-curve (Hybrid, openmpi)<br /> |- <br /> ! acl0-5<br /> ! bs0-5 GigE<br /> ! bs0-5 IB<br /> ! acl0-5<br /> ! bs0-5 GigE<br /> ! bs0-5 IB<br /> ! acl0-5<br /> ! bs0-5 GigE<br /> ! bs0-5 IB<br /> |-<br /> | np X-XX<br /> | 2-20<br /> | 2-48<br /> | 2-48<br /> | 2-12<br /> | 2-48<br /> | 2-48<br /> | 2-20<br /> | 2-48<br /> | 2-48<br /> |}<br /> <br /> What works so far? B = builds, R = runs, W = works<br /> {| class=&quot;wikitable&quot; border=&quot;1&quot;<br /> ! rowspan=&quot;2&quot; | B-builds, R-runs<br /> ! colspan=&quot;4&quot; | area under curve<br /> ! colspan=&quot;4&quot; | GalaxSee (standalone)<br /> |-<br /> ! Serial<br /> ! MPI<br /> ! OpenMP<br /> ! Hybrid<br /> ! Serial<br /> ! MPI<br /> ! OpenMP<br /> ! Hybrid<br /> |-<br /> ! acls<br /> | BRW<br /> | BRW<br /> | BRW<br /> | BRW<br /> | <br /> | BRW<br /> | <br /> |<br /> |-<br /> ! bobsced0<br /> | BRW<br /> | BRW<br /> | BRW<br /> | BRW<br /> |<br /> | BRW<br /> | <br /> |<br /> |-<br /> ! c13<br /> |<br /> |<br /> |<br /> |<br /> |<br /> | BRW<br /> |<br /> |<br /> |-<br /> ! pople<br /> |<br /> |<br /> |<br /> |<br /> |<br /> |<br /> |<br /> |<br /> |-<br /> ! Charlie's laptop<br /> |<br /> | <br /> |<br /> |<br /> |<br /> | BRW<br /> |<br /> |<br /> |}<br /> <br /> To Do<br /> * Fitz/Charlie's message<br /> * Petascale review<br /> * BobSCEd stress test<br /> <br /> Implementations of area under the curve<br /> * Serial<br /> * OpenMP (shared)<br /> * MPI (message passing)<br /> * MPI (hybrid mp and shared)<br /> * OpenMP + MPI (hybrid)<br /> <br /> GalaxSee Goals<br /> * Good piece of code, serves as teaching example for n-body problems in petascale.<br /> * Dials, knobs, etc. in place to easily control how work is distributed when running in parallel.<br /> * Architecture generally supports hybrid model running on large-scale constellations.<br /> * Produces runtime data that enables nice comparisons across multiple resources (scaling, speedup, efficiency).<br /> * Render in BCCD, metaverse, and /dev/null environments.<br /> * Serial version<br /> * Improve performance on math?<br /> <br /> GalaxSee - scale to petascale with MPI and OpenMP hybrid.<br /> * GalaxSee - render in-world and steer from in-world.<br /> * Area under a curve - serial, MPI, and OpenMP implementations.<br /> * OpenMPI - testing, performance.<br /> * Start May 11th<br /> <br /> LittleFe<br /> * Testing<br /> * Documentation <br /> * Touch screen interface<br /> <br /> Notes from May 21, 2009 Review<br /> * Combined Makefiles with defines to build on a particular platform<br /> * Write a driver script for GalaxSee ala the area under the curve script, consider combining<br /> * Schema <br /> ** date, program_name, program_version, style, command line, compute_resource, NP, wall_time <br /> * Document the process from start to finish<br /> * Consider how we might iterate over e.g. number of stars, number of segments, etc.<br /> * Command line option to stat.pl that provides a Torque wrapper for the scripts.<br /> * Lint all code, consistent formatting<br /> * Install latest and greatest Intel compiler in /cluster/bobsced<br /> <br /> == BobSCEd Upgrade ==<br /> Build a new image for BobSCEd:<br /> # One of the Suse versions supported for Gaussian09 on EM64T [v11.1] - Red Hat Enterprise Linux 5.3; SuSE Linux 9.3, 10.3, 11.1; or SuSE Linux Enterprise 10 (see [http://www.gaussian.com/g09_plat.htm G09 platform list]) &lt;-- CentOS 5.3 runs Gaussian binaries for RHEL ok<br /> # Firmware update?<br /> # C3 tools and configuration [v4.0.1]<br /> # Ganglia and configuration [v3.1.2]<br /> # PBS and configuration [v2.3.16]<br /> # /cluster/bobsced local to bs0 <br /> # /cluster/... passed-through to compute nodes<br /> # Large local scratch space on each node<br /> # Gaussian09<br /> # WebMO and configuration [v9.1]<br /> # [[Bobsced Infiniband | Infiniband and configuration]]<br /> # GNU toolchain with OpenMPI and MPICH [GCC v4.4.0], [OpenMPI v1.3.2] [MPICH v1.2.7p1]<br /> # Intel toolchain with OpenMPI and native libraries<br /> # Sage with do-dads (see Charlie)<br /> # Systemimager for the client nodes?<br /> <br /> Installed:<br /> * [[Cluster: New BobSCEd Install Log | New BobSCEd Install Log]]<br /> <br /> Fix the broken nodes.<br /> <br /> == (Old) To Do ==<br /> BCCD Liberation<br /> * v1.1 release - upgrade procedures<br /> <br /> Curriculum Modules <br /> * POVRay<br /> * GROMACS<br /> * Energy and Weather <br /> * Dave's math modules<br /> * Standard format, templates, how-to for V and V<br /> <br /> LittleFe<br /> * Explore machines from first Intel donation ([[intel-lf-server|notes and pictures]])<br /> * Build 4 SCED units<br /> <br /> Infrastructure<br /> * Masa's GROMACS interface on Cairo<br /> * gridgate configuration, Open Science Grid peering<br /> * [[hopperprime|hopper']]<br /> <br /> SC Education <br /> * Scott's homework (see [[sc-education-homework-1|the message]])<br /> <br /> == Current Projects ==<br /> * [[BCCD]]<br /> * [[LittleFe Cluster|LittleFe]]<br /> * [[Folding@Clusters|Folding@Clusters]]<br /> * [[OpenMPI|Benchmarking OpenMPI]]<br /> <br /> == Past Projects ==<br /> * [[Cluster:Big-Fe|Big-FE]]<br /> * [[Cluster:LowLatency|Low Latency Linux Kernel]]<br /> <br /> == General Stuff == <br /> * [[Cluster:Todo|Todo]]<br /> * [[General Cluster Information|General]]<br /> * [[Cluster:Hopper|Hopper]]<br /> * [[Cluster Howto's|Howto's]]<br /> * [[Cluster:Networking|Networking]]<br /> * [[Cluster:2005-11-30 Meeting|2005-11-30 Meeting]]<br /> * [[Cluster:2006-12-12 Meeting|2006-12-12 Meeting]]<br /> * [[Cluster:2006-02-02 Meeting|2006-02-02 Meeting]]<br /> * [[Cluster:2006-03-16 Meeting|2006-03-16 Meeting]]<br /> * [[Cluster:2006-04-06 Meeting|2006-04-06 Meeting]]<br /> * [[Cluster:Node usage|Node usage]]<br /> * [[Cluster:Netgear numbers|Numbers for Netgear switches]]<br /> * [[Cluster:Latex poster creation|Latex Poster Creation]]<br /> * [[Cluster:Bugzilla|Bugzilla Etiquette]]<br /> * [[Cluster:Modules|Modules]]<br /> <br /> == Items Particular to a Specific Cluster ==<br /> * [[ACL Cluster|ACL]]<br /> * [[Athena Cluster|Athena]]<br /> * [[Bazaar Cluster|Bazaar]]<br /> * [[Cairo Cluster|Cairo]]<br /> * [[Bobsced Cluster|Bobsced]]<br /> <br /> == Curriculum Modules ==<br /> * [[Cluster:Curriculum|Curriculum]]<br /> * [[Cluster:Fluid Dynamics|Fluid Dynamics]]<br /> * [[Cluster:Population Ecology|Population Ecology]]<br /> * [[Cluster:GROMACS Web Interface|GROMACS Web Interface]]<br /> * [[Cluster:Wiki|Wiki Life for Academics]]<br /> <br /> == Possible Future Projects ==<br /> * [[Cluster:Realtime Parallel Visualization|Realtime Parallel Visualization]]<br /> <br /> == Archive ==<br /> * TeraGrid '06 (Indianapolis, June 12-15, 2006)<br /> ** [http://www.teragrid.org Conference webpage]<br /> ** [http://www.teragrid.org/events/2006conference/contest_poster.html Student poster guidelines]<br /> ** [[Big-FE-teragrid-abstract|Big-FE abstract]]<br /> <br /> * SIAM Parallel Processing 2006 (San Fransisco, February 22-24, 2006)<br /> ** [http://www.siam.org/meetings/pp06 Conference webpage] <br /> ** [[Cluster:little-fe-siam-pp06-abstract|Little-Fe abstract]]<br /> ** [[Cluster:llk-siam-pp06-abstract|Low Latency Kernal abstract]]<br /> ** Folding@Clusters<br /> ** Best practices for teaching parallel programming to science faculty (Charlie only)<br /> <br /> * [[College Avenue]]</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hopperprime&diff=3867 Hopperprime 2008-01-15T03:36:30Z <p>Skylar: </p> <hr /> <div>= hp TODO =<br /> <br /> Located at &lt;tt&gt;hp.cluster.earlham.edu, 159.28.234.108&lt;/tt&gt;<br /> <br /> Description:<br /> &lt;p&gt;The hardware is an S5000 motherboard with 2 3.2GHz dual-core Xeons<br /> (with hyper-threading enabled, i.e. 8 cores total), 2GB RAM, and two<br /> 250GB SATA drives configured as RAID1. <br /> *Intel's description of the board is here: http://www.intel.com/support/motherboards/server/s5000pal/index.htm<br /> *The manuals and memory specs are in the sna CVS module in /cluster/cvsroot<br /> <br /> Tasks:<br /> * (Complete) System Install [CP] 20070613<br /> * (Complete) sudo [HH] 20070613<br /> * (Complete) Need /usr/src [HH] 20070613<br /> * (Complete) Security Patches (Keep downloaded patches in /root/patches) [HH] 20070615<br /> ** (Complete) Applied FreeBSD-SA-07:02.bin [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:03.ipv6 [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:04.file [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:08.openssl [HH] 20071006<br /> ** (Complete) Applied FreeBSD-SA-07:09.random [HH] 20071130<br /> ** (Complete) Applied FreeBSD-SA-07:10.gtar [HH] 20071130<br /> * (Complete) cvsup [HH] 20070615<br /> ** (Complete) Setup cron job [HH] 20070613<br /> * (Complete) SSH config and banner [HH] 20070613<br /> * (Complete) portupgrade [HH] 20070615<br /> * (Complete) Upgrade X to X 7 (all x-{related_packages}) [HH] 20070615<br /> * (Complete) Screen [NM]<br /> * (Complete) Pine + Pico [NM]<br /> * (Complete) Bash [HH] 20070615<br /> * (Complete) Apache2 /PHP,MySQL,Session [HH] 20070623<br /> * (Complete) [[ntpd]] [NM]<br /> * (Complete) NFS (/Cluster over TCP) [HH] 20070623<br /> * (Complete) Password migration [HH] 20070624<br /> * (Complete) [[MYSQL]] [NM]<br /> * (Complete) [[PortAudit]] [NM]<br /> * (Complete) [[PROFTPD]] [NM]<br /> * (Complete) chkrootkit [NM]<br /> * (Complete) [[Bugzilla]] (over MySQL).[NM]<br /> * (Complete) [[PostgreSQL]]. [NM]<br /> * (Complete) Ganglia [HH] 20070819<br /> * (Complete) SSH Keys from Hopper [HH] 20070819<br /> * (Complete) DNS [HH] 20070819<br /> * (Complete) Cacti [HH]<br /> * (Complete) New NIC Card [multiple ports] [CP]<br /> * (Complete) DHCPD [HH]<br /> * (Complete) Talk [HH]<br /> * (Complete) Kernel tweaks [HH] 20071202<br /> * (Not Installed) NIS [HH]<br /> * (Not Installed) CVS [NM]<br /> * (Complete) [[Hoperrprime:Trac]] [ST]<br /> * (Complete) [[Hopperprime:WebDAV]] [ST]<br /> * (Not Installed) Remote Access Controller card [CP]<br /> * (Not Complete) Final Cut-over (mysql/postgreSQL/Bugzilla) [NM]<br /> * Figure-out what's on admin that should move to hp (so we can abandon admin) [CP]<br /> * Setup NFS over UDP and TCP sharing /cluster (after copying the current /cluster) [CP]</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:WebDAV&diff=5878 Hoperrprime:WebDAV 2008-01-15T03:35:05Z <p>Skylar: </p> <hr /> <div># Build Subversion with Apache support by adding the &lt;tt&gt;WITH_MOD_DAV_SVN=yes WITH_APACHE2_APR=yes&lt;/tt&gt; options to the build.<br /> # Make sure the Apache modules are loaded in &lt;tt&gt;httpd.conf&lt;/tt&gt;:<br /> &lt;pre&gt;<br /> LoadModule dav_svn_module libexec/apache22/mod_dav_svn.so<br /> LoadModule authz_svn_module libexec/apache22/mod_authz_svn.so<br /> &lt;/pre&gt;<br /> # Create a &lt;tt&gt;Includes/svn.conf&lt;/tt&gt; with<br /> <br /> &lt;pre&gt;<br /> &lt;IfModule dav_svn_module&gt;<br /> &lt;Location /svn&gt;<br /> DAV svn<br /> SVNPath /cluster/svnroot<br /> &lt;/Location&gt;<br /> &lt;/IfModule&gt;<br /> <br /> &lt;Location /detail/svnroot&gt;<br /> Order allow,deny<br /> Allow from none<br /> Deny from all<br /> &lt;/Location&gt;<br /> &lt;/pre&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:WebDAV&diff=3774 Hoperrprime:WebDAV 2008-01-15T03:29:22Z <p>Skylar: </p> <hr /> <div># Build Subversion with Apache support by adding the &lt;tt&gt;WITH_MOD_DAV_SVN=yes WITH_APACHE2_APR=yes&lt;/tt&gt; options to the build.<br /> # Make sure the Apache modules are loaded in &lt;tt&gt;httpd.conf&lt;/tt&gt;:<br /> &lt;pre&gt;<br /> LoadModule dav_svn_module libexec/apache22/mod_dav_svn.so<br /> LoadModule authz_svn_module libexec/apache22/mod_authz_svn.so<br /> &lt;/pre&gt;<br /> # Create a &lt;tt&gt;Includes/svn.conf&lt;/tt&gt; with<br /> <br /> &lt;pre&gt;<br /> &lt;IfModule dav_svn_module&gt;<br /> &lt;Location /svn&gt;<br /> DAV svn<br /> SVNPath /cluster/svnroot<br /> &lt;/Location&gt;<br /> &lt;/IfModule&gt;<br /> <br /> &lt;Location /detail/svnroot&gt;<br /> Order allow,deny<br /> Allow from none<br /> Deny from all<br /> &lt;/Location&gt;<br /> &lt;/pre&gt;<br /> # Install WebSVN<br /> # Add an alias for it: &lt;tt&gt;Alias /WebSVN &quot;/usr/local/www/data/WebSVN&quot;&lt;/tt&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:WebDAV&diff=3773 Hoperrprime:WebDAV 2008-01-15T03:28:45Z <p>Skylar: </p> <hr /> <div># Build Subversion with Apache support by adding the &lt;tt&gt;WITH_MOD_DAV_SVN=yes WITH_APACHE2_APR=yes&lt;/tt&gt; options to the build.<br /> # Make sure the Apache modules are loaded in &lt;tt&gt;httpd.conf&lt;/tt&gt;:<br /> &lt;pre&gt;<br /> LoadModule dav_svn_module libexec/apache22/mod_dav_svn.so<br /> LoadModule authz_svn_module libexec/apache22/mod_authz_svn.so<br /> &lt;/pre&gt;<br /> # Create a &lt;tt&gt;Includes/svn.conf&lt;/tt&gt; with<br /> <br /> &lt;pre&gt;<br /> &lt;IfModule dav_svn_module&gt;<br /> &lt;Location /svn&gt;<br /> DAV svn<br /> SVNPath /cluster/svnroot<br /> &lt;/Location&gt;<br /> &lt;/IfModule&gt;<br /> <br /> &lt;Location /detail/svnroot&gt;<br /> Order allow,deny<br /> Allow from none<br /> Deny from all<br /> &lt;/Location&gt;<br /> &lt;/pre&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:Trac&diff=5877 Hoperrprime:Trac 2008-01-15T03:18:55Z <p>Skylar: </p> <hr /> <div># Subversion<br /> ## Special compile for apr-svn w/ &lt;tt&gt;APR_UTIL_WITH_BERKELEY_DB=yes&lt;/tt&gt;<br /> ## portinstall subversion<br /> ## svnadmin hotcopy to move Subversion over NFS<br /> # Install mod_python-3.3.1 from ports<br /> # Install trac-0.10.4 from ports (made sure Postgres support was enabled)<br /> # Loaded trac_debian-cluster Postgres database from hopper with pgdump and psql<br /> # Added &lt;tt&gt;Includes/trac.conf&lt;/tt&gt; to apache configuration, based on TracModPython documentation.<br /> <br /> &lt;pre&gt;<br /> Alias &quot;/trac&quot; &quot;/cluster/trac&quot;<br /> <br /> &lt;Location /trac&gt;<br /> SetHandler mod_python<br /> PythonHandler trac.web.modpython_frontend<br /> PythonOption TracEnvParentDir /cluster/trac<br /> PythonOption TracUriRoot /trac<br /> &lt;/Location&gt;<br /> <br /> &lt;Location &quot;/trac/debian-cluster/login&quot;&gt;<br /> AuthType Basic<br /> AuthName &quot;Trac&quot;<br /> AuthUserFile /cluster/trac/htpasswd<br /> require valid-user<br /> &lt;/Location&gt;<br /> &lt;/pre&gt;</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:Trac&diff=3772 Hoperrprime:Trac 2008-01-15T03:18:01Z <p>Skylar: </p> <hr /> <div># Subversion<br /> ## Special compile for apr-svn w/ &lt;tt&gt;APR_UTIL_WITH_BERKELEY_DB=yes&lt;/tt&gt;<br /> ## portinstall subversion<br /> ## svnadmin hotcopy to move Subversion over NFS<br /> # Install mod_python-3.3.1 from ports<br /> # Install trac-0.10.4 from ports (made sure Postgres support was enabled)<br /> # Loaded trac_debian-cluster Postgres database from hopper with pgdump and psql<br /> # Added &lt;tt&gt;Includes/trac.conf&lt;/tt&gt; to apache configuration, based on TracModPython documentation.</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hopperprime&diff=3775 Hopperprime 2008-01-15T02:51:36Z <p>Skylar: </p> <hr /> <div>= hp TODO =<br /> <br /> Located at &lt;tt&gt;hp.cluster.earlham.edu, 159.28.234.108&lt;/tt&gt;<br /> <br /> Description:<br /> &lt;p&gt;The hardware is an S5000 motherboard with 2 3.2GHz dual-core Xeons<br /> (with hyper-threading enabled, i.e. 8 cores total), 2GB RAM, and two<br /> 250GB SATA drives configured as RAID1. <br /> *Intel's description of the board is here: http://www.intel.com/support/motherboards/server/s5000pal/index.htm<br /> *The manuals and memory specs are in the sna CVS module in /cluster/cvsroot<br /> <br /> Tasks:<br /> * (Complete) System Install [CP] 20070613<br /> * (Complete) sudo [HH] 20070613<br /> * (Complete) Need /usr/src [HH] 20070613<br /> * (Complete) Security Patches (Keep downloaded patches in /root/patches) [HH] 20070615<br /> ** (Complete) Applied FreeBSD-SA-07:02.bin [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:03.ipv6 [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:04.file [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:08.openssl [HH] 20071006<br /> ** (Complete) Applied FreeBSD-SA-07:09.random [HH] 20071130<br /> ** (Complete) Applied FreeBSD-SA-07:10.gtar [HH] 20071130<br /> * (Complete) cvsup [HH] 20070615<br /> ** (Complete) Setup cron job [HH] 20070613<br /> * (Complete) SSH config and banner [HH] 20070613<br /> * (Complete) portupgrade [HH] 20070615<br /> * (Complete) Upgrade X to X 7 (all x-{related_packages}) [HH] 20070615<br /> * (Complete) Screen [NM]<br /> * (Complete) Pine + Pico [NM]<br /> * (Complete) Bash [HH] 20070615<br /> * (Complete) Apache2 /PHP,MySQL,Session [HH] 20070623<br /> * (Complete) [[ntpd]] [NM]<br /> * (Complete) NFS (/Cluster over TCP) [HH] 20070623<br /> * (Complete) Password migration [HH] 20070624<br /> * (Complete) [[MYSQL]] [NM]<br /> * (Complete) [[PortAudit]] [NM]<br /> * (Complete) [[PROFTPD]] [NM]<br /> * (Complete) chkrootkit [NM]<br /> * (Complete) [[Bugzilla]] (over MySQL).[NM]<br /> * (Complete) [[PostgreSQL]]. [NM]<br /> * (Complete) Ganglia [HH] 20070819<br /> * (Complete) SSH Keys from Hopper [HH] 20070819<br /> * (Complete) DNS [HH] 20070819<br /> * (Complete) Cacti [HH]<br /> * (Complete) New NIC Card [multiple ports] [CP]<br /> * (Complete) DHCPD [HH]<br /> * (Complete) Talk [HH]<br /> * (Complete) Kernel tweaks [HH] 20071202<br /> * (Not Installed) NIS [HH]<br /> * (Not Installed) CVS [NM]<br /> * (Not Complete) [[Hoperrprime:Trac]] [ST]<br /> * (Not Complete) [[Hopperprime:WebDAV]] [ST]<br /> * (Not Installed) Remote Access Controller card [CP]<br /> * (Not Complete) Final Cut-over (mysql/postgreSQL/Bugzilla) [NM]<br /> * Figure-out what's on admin that should move to hp (so we can abandon admin) [CP]<br /> * Setup NFS over UDP and TCP sharing /cluster (after copying the current /cluster) [CP]</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hoperrprime:Trac&diff=3771 Hoperrprime:Trac 2008-01-15T02:50:49Z <p>Skylar: </p> <hr /> <div># Subversion<br /> ## Special compile for apr-svn w/ &lt;tt&gt;APR_UTIL_WITH_BERKELEY_DB=yes&lt;/tt&gt;<br /> ## portinstall subversion<br /> # Install mod_python-3.3.1 from ports<br /> # Install trac-0.10.4 from ports</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hopperprime&diff=3770 Hopperprime 2008-01-15T02:46:10Z <p>Skylar: </p> <hr /> <div>= hp TODO =<br /> <br /> Located at &lt;tt&gt;hp.cluster.earlham.edu, 159.28.234.108&lt;/tt&gt;<br /> <br /> Description:<br /> &lt;p&gt;The hardware is an S5000 motherboard with 2 3.2GHz dual-core Xeons<br /> (with hyper-threading enabled, i.e. 8 cores total), 2GB RAM, and two<br /> 250GB SATA drives configured as RAID1. <br /> *Intel's description of the board is here: http://www.intel.com/support/motherboards/server/s5000pal/index.htm<br /> *The manuals and memory specs are in the sna CVS module in /cluster/cvsroot<br /> <br /> Tasks:<br /> * (Complete) System Install [CP] 20070613<br /> * (Complete) sudo [HH] 20070613<br /> * (Complete) Need /usr/src [HH] 20070613<br /> * (Complete) Security Patches (Keep downloaded patches in /root/patches) [HH] 20070615<br /> ** (Complete) Applied FreeBSD-SA-07:02.bin [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:03.ipv6 [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:04.file [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:08.openssl [HH] 20071006<br /> ** (Complete) Applied FreeBSD-SA-07:09.random [HH] 20071130<br /> ** (Complete) Applied FreeBSD-SA-07:10.gtar [HH] 20071130<br /> * (Complete) cvsup [HH] 20070615<br /> ** (Complete) Setup cron job [HH] 20070613<br /> * (Complete) SSH config and banner [HH] 20070613<br /> * (Complete) portupgrade [HH] 20070615<br /> * (Complete) Upgrade X to X 7 (all x-{related_packages}) [HH] 20070615<br /> * (Complete) Screen [NM]<br /> * (Complete) Pine + Pico [NM]<br /> * (Complete) Bash [HH] 20070615<br /> * (Complete) Apache2 /PHP,MySQL,Session [HH] 20070623<br /> * (Complete) [[ntpd]] [NM]<br /> * (Complete) NFS (/Cluster over TCP) [HH] 20070623<br /> * (Complete) Password migration [HH] 20070624<br /> * (Complete) [[MYSQL]] [NM]<br /> * (Complete) [[PortAudit]] [NM]<br /> * (Complete) [[PROFTPD]] [NM]<br /> * (Complete) chkrootkit [NM]<br /> * (Complete) [[Bugzilla]] (over MySQL).[NM]<br /> * (Complete) [[PostgreSQL]]. [NM]<br /> * (Complete) Ganglia [HH] 20070819<br /> * (Complete) SSH Keys from Hopper [HH] 20070819<br /> * (Complete) DNS [HH] 20070819<br /> * (Complete) Cacti [HH]<br /> * (Complete) New NIC Card [multiple ports] [CP]<br /> * (Complete) DHCPD [HH]<br /> * (Complete) Talk [HH]<br /> * (Complete) Kernel tweaks [HH] 20071202<br /> * (Not Installed) NIS [HH]<br /> * (Not Installed) CVS [NM]<br /> * (Not Complete) [[Hoperrprime:Trac]] [ST]<br /> * (Not Installed) Remote Access Controller card [CP]<br /> * (Not Complete) Final Cut-over (mysql/postgreSQL/Bugzilla) [NM]<br /> * Figure-out what's on admin that should move to hp (so we can abandon admin) [CP]<br /> * Setup NFS over UDP and TCP sharing /cluster (after copying the current /cluster) [CP]</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Hopperprime&diff=3769 Hopperprime 2008-01-15T02:45:51Z <p>Skylar: </p> <hr /> <div>= hp TODO =<br /> <br /> Located at &lt;tt&gt;hp.cluster.earlham.edu, 159.28.234.108&lt;/tt&gt;<br /> <br /> Description:<br /> &lt;p&gt;The hardware is an S5000 motherboard with 2 3.2GHz dual-core Xeons<br /> (with hyper-threading enabled, i.e. 8 cores total), 2GB RAM, and two<br /> 250GB SATA drives configured as RAID1. <br /> *Intel's description of the board is here: http://www.intel.com/support/motherboards/server/s5000pal/index.htm<br /> *The manuals and memory specs are in the sna CVS module in /cluster/cvsroot<br /> <br /> Tasks:<br /> * (Complete) System Install [CP] 20070613<br /> * (Complete) sudo [HH] 20070613<br /> * (Complete) Need /usr/src [HH] 20070613<br /> * (Complete) Security Patches (Keep downloaded patches in /root/patches) [HH] 20070615<br /> ** (Complete) Applied FreeBSD-SA-07:02.bin [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:03.ipv6 [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:04.file [HH] 20070613<br /> ** (Complete) Applied FreeBSD-SA-07:08.openssl [HH] 20071006<br /> ** (Complete) Applied FreeBSD-SA-07:09.random [HH] 20071130<br /> ** (Complete) Applied FreeBSD-SA-07:10.gtar [HH] 20071130<br /> * (Complete) cvsup [HH] 20070615<br /> ** (Complete) Setup cron job [HH] 20070613<br /> * (Complete) SSH config and banner [HH] 20070613<br /> * (Complete) portupgrade [HH] 20070615<br /> * (Complete) Upgrade X to X 7 (all x-{related_packages}) [HH] 20070615<br /> * (Complete) Screen [NM]<br /> * (Complete) Pine + Pico [NM]<br /> * (Complete) Bash [HH] 20070615<br /> * (Complete) Apache2 /PHP,MySQL,Session [HH] 20070623<br /> * (Complete) [[ntpd]] [NM]<br /> * (Complete) NFS (/Cluster over TCP) [HH] 20070623<br /> * (Complete) Password migration [HH] 20070624<br /> * (Complete) [[MYSQL]] [NM]<br /> * (Complete) [[PortAudit]] [NM]<br /> * (Complete) [[PROFTPD]] [NM]<br /> * (Complete) chkrootkit [NM]<br /> * (Complete) [[Bugzilla]] (over MySQL).[NM]<br /> * (Complete) [[PostgreSQL]]. [NM]<br /> * (Complete) Ganglia [HH] 20070819<br /> * (Complete) SSH Keys from Hopper [HH] 20070819<br /> * (Complete) DNS [HH] 20070819<br /> * (Complete) Cacti [HH]<br /> * (Complete) New NIC Card [multiple ports] [CP]<br /> * (Complete) DHCPD [HH]<br /> * (Complete) Talk [HH]<br /> * (Complete) Kernel tweaks [HH] 20071202<br /> * (Not Installed) NIS [HH]<br /> * (Not Installed) CVS [NM]<br /> * (Not Complete) Trac [ST]<br /> * (Not Installed) Remote Access Controller card [CP]<br /> * (Not Complete) Final Cut-over (mysql/postgreSQL/Bugzilla) [NM]<br /> * Figure-out what's on admin that should move to hp (so we can abandon admin) [CP]<br /> * Setup NFS over UDP and TCP sharing /cluster (after copying the current /cluster) [CP]</div> Skylar https://wiki.cs.earlham.edu/index.php?title=File:New_fiber_plan.jpg&diff=5538 File:New fiber plan.jpg 2006-04-24T05:30:42Z <p>Skylar: Post April '06 fiber plan</p> <hr /> <div>Post April '06 fiber plan</div> Skylar https://wiki.cs.earlham.edu/index.php?title=File:Old_fiber_plan.jpg&diff=5439 File:Old fiber plan.jpg 2006-04-24T05:27:12Z <p>Skylar: Pre-April '06 Fiber Plan</p> <hr /> <div>Pre-April '06 Fiber Plan</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:2006-04-06_Meeting&diff=5532 Cluster:2006-04-06 Meeting 2006-04-06T16:50:02Z <p>Skylar: </p> <hr /> <div>* Look at hopeless (Toby, Skylar)<br /> * Finish up bigfe (JoshM)<br /> * Send pointer to TomM's folks wrt BCCD download (Skylar)<br /> * Reconfigure weatherduck db (Skylar)<br /> * Look at apt for BCCD (Toby)<br /> * TeraGrid<br /> * Backup (Skylar)<br /> ** Get magazine<br /> ** Fix admin AMANDA backups<br /> * Backup c10/c11 (Toby)<br /> * Start using Bugzilla for BCCD liberation (Skylar, Kevin)</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:2006-04-06_Meeting&diff=1538 Cluster:2006-04-06 Meeting 2006-04-06T16:41:19Z <p>Skylar: </p> <hr /> <div>* Look at hopeless (Toby, Skylar)<br /> * Finish up bigfe (JoshM)<br /> * Send pointer to TomM's folks wrt BCCD download (Skylar)<br /> * Reconfigure weatherduck db (Skylar)<br /> * Look at apt for BCCD (Toby)<br /> * TeraGrid<br /> * Backup (Skylar)<br /> ** Get magazine<br /> ** Fix admin AMANDA backups<br /> * Backup c10/c11 (Toby)</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:2006-04-06_Meeting&diff=1537 Cluster:2006-04-06 Meeting 2006-04-06T16:36:16Z <p>Skylar: </p> <hr /> <div>* Look at hopeless (Toby, Skylar)<br /> * Finish up bigfe (JoshM)<br /> * Send pointer to TomM's folks wrt BCCD download (Skylar)<br /> * Reconfigure weatherduck db (Skylar)<br /> * Look at apt for BCCD (Toby)<br /> * TeraGrid<br /> * Backup (Skylar)<br /> ** Get magazine<br /> ** Fix admin AMANDA backups</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:2006-04-06_Meeting&diff=1536 Cluster:2006-04-06 Meeting 2006-04-06T16:29:03Z <p>Skylar: </p> <hr /> <div>* Look at hopeless (Toby, Skylar)<br /> * Finish up bigfe (JoshM)<br /> * Send pointer to TomM's folks wrt BCCD download (Skylar)<br /> * Reconfigure weatherduck db (Skylar)<br /> * Look at apt for BCCD (Toby)</div> Skylar https://wiki.cs.earlham.edu/index.php?title=Cluster:2006-04-06_Meeting&diff=1535 Cluster:2006-04-06 Meeting 2006-04-06T16:24:21Z <p>Skylar: </p> <hr /> <div>* Look at hopeless (Toby, Skylar)<br /> * Finish up bigfe (JoshM)<br /> * Send pointer to TomM's folks wrt BCCD download (Skylar)<br /> * Reconfigure weatherduck db (Skylar)</div> Skylar