Cluster 0: Difference between revisions

From DISI
Jump to navigation Jump to search
No edit summary
No edit summary
 
(11 intermediate revisions by one other user not shown)
Line 1: Line 1:
This page is about our legacy cluster at [[Mission Bay]].  We also have a [[Cluster 1]] at [[University of Toronto]] and a new cluster [[Cluster 2]] at [[UCSF]].
This page is about our legacy cluster at [[Mission Bay]].  Most users should use [[Cluster 2]].


= Getting started on the cluster =
= Status =
 
Cluster 0 is a legacy cluster and will disappear entirely soon, we hope. It predates many of the policies below. We are trying to get people off Cluster 0 asap and on to Cluster 2.
* 1. request an account from Therese or John.
* 2. Your home is on /raid1/people/<your_id>/. This area is backed up and is for important persistent files.
* 3. You should run docking jobs and other intense calculations in ~/work/, which Therese will set up for you and is generally not your home directory.
* 4. You should keep static data (e.g. crystallography data, results of published papers) in ~/store/ which is generally not your home directory.
* 5. Lab guests get 100GB in each of these areas, and lab members get 500GB. Ask if you need more.
* 6. If you go over your limit, you get emails for 2 weeks, then we impose a hard limit if you have not solved your overage.
* 7. You can choose bash or tcsh to be your default shell. We don't care. Everything should work equally well with both.
* 8. There is a special kind of static data, databases, for which you may request space. They will go in /nfs/db/<db_name>/. e.g. /nfs/db/zinc/ and /nfs/db/dude/ and /nfs/db/pdb and so on.
* 9. Please run large docking jobs on /nfs/work and not on /nfs/store or /nfs/home. When you publish a paper, please delete what you can, compress the rest, and move it to /store/. Do not leave it on /work/ if you are no longer using it actively.
* 10. Set up your account so that you can log in all across the cluster without a password. For instructions on how to securely generate ssh keys go here: http://wiki.uoft.bkslab.org/index.php/How_to_generate_ssh_keys_securely
* 11. Software lives in /nfs/software/. All our machines are 64 bit Centos 6.3 unless otherwise indicated.
* 12. Python 2.7 and 3.0 are installed. We currently recommend 2.7 because of library availability, but that may change soon. (Aug 2012)
* 13. If you use tcsh, copy .login and .cshrc from ~jji/  ; If you use bash, copy .bash_profile from ~jji/
 
 
* 1. cp /nfs/software/labenv/defaults.cshrc .cshrc
Note: if you are still in San Francisco, the path is /raid3/software/labenv/defaults.cshrc
If you use bash or another shell, please see the Sysadmin.
* 2. Customize this file if you like.
* 3. Check out your own copy of dockenv, dock, sea, if you like.
By default you use the standard lab software.
* 4. Logout / login or source ~/.cshrc
* 5. You are now ready to use all the lab software, including docking.


= Priorities and Policies =
* [[Lab Security Policy]]
* [[Disk space policy]]
* [[Backups]] policy.
* [[Portal system]] for off-site ssh cluster access.
* Get a [[Cluster 0 account]] and get started


= Physical machine summary =  
= Physical machine summary =  
<pre>
<pre>
256 Intel Xeon E5430 2.66Ghz cores (8core)
256 Intel Xeon E5430 2.66Ghz cores (8core)
Line 38: Line 20:
24  AMD 6164HE 1.7ghz cores      (dl165g7)
24  AMD 6164HE 1.7ghz cores      (dl165g7)
</pre>
</pre>
* Each node has 1GB memory/cpu-core or better
* Stored in Racks 0,1,2,3,4,5,6,7
* 36TB of RAID10 storage available to cluster, hosted among 4 dedicated NFS servers
* 10 Support/Infrastructure servers (e.g. databases)


* Four node racks and Two server racks in BH-101
= Database machines =
* Two node racks and One server rack in N108
* Scratch is a general purpose mysql server
* 36TB of RAID10 storage available to cluster, hosted among 8 dedicated servers
* zincdb1 is mysql for zinc (master).  zincdb4 is a slave focusing on structure smiles/smarts search.  zincdb6 is a slave focusing on dude and decoys.
* Each node has 1GB memory/cpu-core or better
 
= Detailed Machine Breakdown =
Warning: dated.
<pre>
Totals:
 
Nodes 177
Servers real 25
Servers virt 18
Desktops real 20
Desktops virtual 18
Desktops insturment 7
Printers 3
Laptops 7
 
Nodes:
    N108:
        Node rack 3:
            34 HP DL140G1
 
        Node rack 4:
            19 Microway nodes
 
    BH101:
        Node rack 1:
            40 HP DL140G2
 
        Node rack 2:
            15 HP DL140G2
            25 HP DL140G1
 
        Node rack 5:
            7 HP DL145G2
            20 HP DL160G5
 
        Node rack 6:
            17 HP DL160G5
            1  HP DL160G5 ( node-6-20, Carchia development machine )
 
Servers:
    N108:
        Server Rack 2:
            Korn - FC5 - Primarily for Daylight fingerprint serving
            NIS  - Centos 5 - Core NIS server.
            zincdb2 - FC7 - Slave zincdb server
            nfshead1 - FC5 - /raid1, /raid2 NFS server
            aaa2 - Centos 5.4 - Backup DNS/LDAP/Logging server
            offspring - FC6 - /raid5 NFS server
            scratch - Centos 5 - Misc MySQL server
            sgehead2 - Secondary SGE submission node
            clash - Centos 5.4 - /raid0 NFS server
 
    BH101:
        Server Rack 1:
            ppilot - Centos 5.3 - Pipeline pilot server
            aaa1 - Centos 5.4 - PRimary DNS/LDAP/Logging server
            wilco - FC7 - DOCK BLASTER webserver
            mon - FC6 - Nagios, OTRS, monitoring system
            zincdb1 - FC7 - Master zincdb server
            marilyn2 - FC4 - WIKI server, misc db server, mol img server
            sgehead1 - Centos 5.4 - Primary SGE headnode
            sgemaster - Centos 5.4 - SGE Masster server
            sea1 - Centos 5.4 - (8core) Standalone SEA server
            sea2 - Centos 5.4 - (2core) Standalone SEA server
            vmware2 - ESXi 4.0 - ESXi server for windows workstations
            tools - Centos 5.3 - General tools server
 
        Server Rack 3:
            nfshead2 - Centos 5.4 - raid3, raid6, raid7 NFS server
            zincdata1 - Centos 5.4 - prototype ZINC server
            backup - Centos 5.4 - Arkeia backup server
            vmware1 - ESXi 4.0 - ESXi servers
 
    vmware1 virtual:
        bks150
        bks151
        bks153
        bks154
        bkswinsrv - Win server 2003 - PDC
        dock - FC1 - Legacy DOCK webserver
        filemaker - Win XP - Filemaker server
        ftp.bkslab.org - Centos 5.4 - general file transfer server
        hamlet - Centos 5.3 - DOCK CVS server
        hiers - Centos 5.3 - Server for jji
        hkl2000.bkslab.org - Centos 5.4 - HKL2000 licensed server
        installer.bkslab.org - Centos 5.4 - Cluster kickstart, dhcpd server, backups of network switch configs in /tftpboot
        jerry - RHEL WS 4 - Oracle server
        kmfdm - Win XP - General fileserver, IRIS/Base server
        mailman.bkslab.org - Centos 5.4 - Lab mailing list server
        svn.bkslab.org - Centos 5.4 - SVN http/svnserve read only server
        tripos - RH 7.3 - Tripos legacy flexlm server
        www.bkslab.org - Centos 5.4 - Lab webserver + supporting mysql
 
    vmware2 virtual ( all winxp ):
        bet
        carchia
        claggner
        dahlia
        eidamo
        hfan
        jaytung
        jens
        jji
        jkarpiak
        jpmotion
        keiser
        knguyen
        kolb
        merski
        mysinger
        pascal
        shoichet
 
Desktops:
    Wetlab user:
        munly - linux - hlin
        clutch - linux - merski
        fourier - linux - rafaela
        pogue - winxp - allie
        meatpuppet - linux - dahlia
        cake - linux - bet
        daft - linux - ryan
        berry - linux - sarah
        roton - winxp - melody
        godspeed - linux - magdelena
 
    Wetlab insturment:
        nin - winxp - spec
        breeder - winxp - spec
        fc - winxp - BC flow cyto
        dls - winxp - Dynapro DLS
        manuchao - winxp - scanner
        tool - winxp - CD spec
        itc - winxp - ITC


* 22 User workstations
    Drylab user:
* 30 Support/Infrastructure servers (e.g. databases)
        newelvis - linux - jens
* 6 Windows Laptops
        elvis - linux - oliv
* 2 VMWARE servers hosting virtual desktops and servers that do not require dedicated hardware
        styx - linux - keiser
        marillion - linux - laggner
        kosh - linux - mysinger
        rage - linux - kolb
        beethoven - linux - grocklin
        zeroth - linux - kong
        x - linux - jji


OK, OK, the laptops and the workstations aren't technically part of the cluster.
    Printers:
        prince
        graupel
        marissa


[[About our cluster]]
    Misc:
        blaise - linux - pascal


[[Server rack 1]]
Laptops:
    brian
    sarah
    presentation
    pascal
    john pc
    john mac
    spare


[[Server rack 2]]
</pre>




[[About our cluster]]
[[Category:Cluster]]
[[Category:Cluster]]
[[Category:Internal]]
[[Category:Internal]]
[[Category:UCSF]]
[[Category:Sysadmin]]

Latest revision as of 00:10, 9 April 2017

This page is about our legacy cluster at Mission Bay. Most users should use Cluster 2.

Status

Cluster 0 is a legacy cluster and will disappear entirely soon, we hope. It predates many of the policies below. We are trying to get people off Cluster 0 asap and on to Cluster 2.

Priorities and Policies

Physical machine summary

256 Intel Xeon E5430 2.66Ghz cores (8core)
118 Intel Xeon 3.0ghz cores        (dl140g1)
106 Intel Xeon 2.8ghz cores        (dl140g2)
38  Intel Xeon 2.4ghz cores        (microway)
32   AMD Opteron 275 2.2ghz cores  (dl145g2)
24   AMD 6164HE 1.7ghz cores       (dl165g7)
  • Each node has 1GB memory/cpu-core or better
  • Stored in Racks 0,1,2,3,4,5,6,7
  • 36TB of RAID10 storage available to cluster, hosted among 4 dedicated NFS servers
  • 10 Support/Infrastructure servers (e.g. databases)

Database machines

  • Scratch is a general purpose mysql server
  • zincdb1 is mysql for zinc (master). zincdb4 is a slave focusing on structure smiles/smarts search. zincdb6 is a slave focusing on dude and decoys.

Detailed Machine Breakdown

Warning: dated.

Totals:

Nodes 177
Servers real 25
Servers virt 18
Desktops real 20
Desktops virtual 18
Desktops insturment 7
Printers 3
Laptops 7

Nodes:
    N108:
        Node rack 3:
            34 HP DL140G1

        Node rack 4:
            19 Microway nodes

    BH101:
        Node rack 1:
            40 HP DL140G2

        Node rack 2:
            15 HP DL140G2
            25 HP DL140G1

        Node rack 5:
            7 HP DL145G2
            20 HP DL160G5

        Node rack 6:
            17 HP DL160G5
            1  HP DL160G5 ( node-6-20, Carchia development machine )

Servers:
    N108:
        Server Rack 2:
            Korn - FC5 - Primarily for Daylight fingerprint serving
            NIS  - Centos 5 - Core NIS server.
            zincdb2 - FC7 - Slave zincdb server
            nfshead1 - FC5 - /raid1, /raid2 NFS server
            aaa2 - Centos 5.4 - Backup DNS/LDAP/Logging server
            offspring - FC6 - /raid5 NFS server
            scratch - Centos 5 - Misc MySQL server
            sgehead2 - Secondary SGE submission node
            clash - Centos 5.4 - /raid0 NFS server

    BH101:
        Server Rack 1:
            ppilot - Centos 5.3 - Pipeline pilot server
            aaa1 - Centos 5.4 - PRimary DNS/LDAP/Logging server
            wilco - FC7 - DOCK BLASTER webserver
            mon - FC6 - Nagios, OTRS, monitoring system
            zincdb1 - FC7 - Master zincdb server
            marilyn2 - FC4 - WIKI server, misc db server, mol img server
            sgehead1 - Centos 5.4 - Primary SGE headnode
            sgemaster - Centos 5.4 - SGE Masster server
            sea1 - Centos 5.4 - (8core) Standalone SEA server
            sea2 - Centos 5.4 - (2core) Standalone SEA server
            vmware2 - ESXi 4.0 - ESXi server for windows workstations
            tools - Centos 5.3 - General tools server

        Server Rack 3:
            nfshead2 - Centos 5.4 - raid3, raid6, raid7 NFS server
            zincdata1 - Centos 5.4 - prototype ZINC server
            backup - Centos 5.4 - Arkeia backup server
            vmware1 - ESXi 4.0 - ESXi servers

    vmware1 virtual:
        bks150
        bks151
        bks153
        bks154
        bkswinsrv - Win server 2003 - PDC 
        dock - FC1 - Legacy DOCK webserver
        filemaker - Win XP - Filemaker server
        ftp.bkslab.org - Centos 5.4 - general file transfer server
        hamlet - Centos 5.3 - DOCK CVS server
        hiers - Centos 5.3 - Server for jji 
        hkl2000.bkslab.org - Centos 5.4 - HKL2000 licensed server
        installer.bkslab.org - Centos 5.4 - Cluster kickstart, dhcpd server, backups of network switch configs in /tftpboot
        jerry - RHEL WS 4 - Oracle server
        kmfdm - Win XP - General fileserver, IRIS/Base server
        mailman.bkslab.org - Centos 5.4 - Lab mailing list server
        svn.bkslab.org - Centos 5.4 - SVN http/svnserve read only server
        tripos - RH 7.3 - Tripos legacy flexlm server
        www.bkslab.org - Centos 5.4 - Lab webserver + supporting mysql

    vmware2 virtual ( all winxp ):
        bet
        carchia
        claggner
        dahlia
        eidamo
        hfan
        jaytung
        jens
        jji
        jkarpiak
        jpmotion
        keiser
        knguyen
        kolb
        merski
        mysinger
        pascal
        shoichet

Desktops:
    Wetlab user:
        munly - linux - hlin
        clutch - linux - merski
        fourier - linux - rafaela
        pogue - winxp - allie
        meatpuppet - linux - dahlia
        cake - linux - bet
        daft - linux - ryan
        berry - linux - sarah
        roton - winxp - melody
        godspeed - linux - magdelena

    Wetlab insturment:
        nin - winxp - spec
        breeder - winxp - spec
        fc - winxp - BC flow cyto
        dls - winxp - Dynapro DLS
        manuchao - winxp - scanner
        tool - winxp - CD spec
        itc - winxp - ITC

    Drylab user:
        newelvis - linux - jens
        elvis - linux - oliv
        styx - linux - keiser
        marillion - linux - laggner
        kosh - linux - mysinger
        rage - linux - kolb
        beethoven - linux - grocklin
        zeroth - linux - kong
        x - linux - jji

    Printers:
        prince
        graupel
        marissa

    Misc:
        blaise - linux - pascal

Laptops:
    brian
    sarah
    presentation
    pascal
    john pc
    john mac
    spare


About our cluster