Equipment/Lattner: Difference between revisions

From London Hackspace Wiki
m (Formatting issues and links.)
Line 37: Line 37:


* 4 1TB drives connected directly to the motherboard SATA ports.
* 4 1TB drives connected directly to the motherboard SATA ports.
* System is using ZFS in a RAIDZ2 (Double-Parity) RAID providing 1800GiB of usable storage.  
* System is using ZFS in a RAIDZ2 (Double-Parity) RAID providing 1800GiB of usable storage:
* zpool was created using device-aware /dev/disk/by-id rather than generic /dev/sda,sdb,sdc,sdd enumeration to help facilitate disk replacements when a drive goes down:
    pool: chris
** HOW?  By labeling the outside of the drive-tray with the disk serial number, we can easily reconcile the serial number of a failed/failing drive with one that is in the array.
  state: ONLINE
    scan: none requested
  config:
 
  NAME                                            STATE    READ WRITE CKSUM
  chris                                          ONLINE      0    0    0
    raidz2-0                                      ONLINE      0    0    0
      ata-Hitachi_HUA721010KLA330_GTA060PBK2ZY1F  ONLINE      0    0    0
      ata-Hitachi_HUA721010KLA330_GTA060PBKEVKZF  ONLINE      0    0    0
      ata-Hitachi_HUA721010KLA330_GTA060PBKEWNTF  ONLINE      0    0    0
      ata-Hitachi_HUA721010KLA330_GTA060PBKEWX4F ONLINE      0    0    0
 
** zpool was created using device-aware /dev/disk/by-id rather than generic /dev/sda,sdb,sdc,sdd enumeration to help facilitate disk replacements when a drive goes down. And by labeling the outside of the drive-tray with the disk serial number, we can easily reconcile the serial number of a failed/failing drive with one that is in the array.
* Boot drive is simply a single SSD - it shouldn't hold anything precious nor be used as a caching device as there is sufficient system memory and ZFS can likely manage caching automatically.
* Boot drive is simply a single SSD - it shouldn't hold anything precious nor be used as a caching device as there is sufficient system memory and ZFS can likely manage caching automatically.

Revision as of 14:20, 8 July 2016

Lattner
Hackspace Unknown.png
Model SuperMicro X7DBU
Sub-category Networking
Status Good working order
Training requirement no
Origin mixed
Location Basement rack
Maintainers Sysadmin team

Lattner is a general-purpose file server provided to help with general hackspace storage stuff. Note that no warranty is given or implied, but we will try our best to keep it reliably online and usable.

The system was named after the computer scientist Chris Lattner. Lattner is a software developer, best known as the main author of LLVM and related projects, such as the compiler Clang and the programming language Swift.

  • 1U Rackmount server with 4 x 1000GB 3.5" SATA drives and one 2.5" 32GB Boot SSD
  • Debian 8
  • Connected to Equipment/Cisco1 via port 31 & 32 in a 2 port LACP channel bond.
  • Runs Samba (workgroup 'LHS') and has 2 shares:
    • space, which is read/write to everyone on the local network
    • managed which contains useful things that are of interest to hackspace members.
  • shares /home over nfs, currently used by lamarr at least.

Motherboard and Memory

Lattner has 32GB of ECC DRAM and is powered by a SuperMicro X7DBU motherboard. Possible management via IPMI is available via an off-board card that is currently NOT installed.

Disk Configuration

  • 4 1TB drives connected directly to the motherboard SATA ports.
  • System is using ZFS in a RAIDZ2 (Double-Parity) RAID providing 1800GiB of usable storage:
   pool: chris
  state: ONLINE
   scan: none requested
 config:
 	NAME                                            STATE     READ WRITE CKSUM
 	chris                                           ONLINE       0     0     0
 	  raidz2-0                                      ONLINE       0     0     0
 	    ata-Hitachi_HUA721010KLA330_GTA060PBK2ZY1F  ONLINE       0     0     0
 	    ata-Hitachi_HUA721010KLA330_GTA060PBKEVKZF  ONLINE       0     0     0
 	    ata-Hitachi_HUA721010KLA330_GTA060PBKEWNTF  ONLINE       0     0     0
 	    ata-Hitachi_HUA721010KLA330_GTA060PBKEWX4F  ONLINE       0     0     0
    • zpool was created using device-aware /dev/disk/by-id rather than generic /dev/sda,sdb,sdc,sdd enumeration to help facilitate disk replacements when a drive goes down. And by labeling the outside of the drive-tray with the disk serial number, we can easily reconcile the serial number of a failed/failing drive with one that is in the array.
  • Boot drive is simply a single SSD - it shouldn't hold anything precious nor be used as a caching device as there is sufficient system memory and ZFS can likely manage caching automatically.