Tembo cluster

tembo.cs is a "paper" cluster. User documentation for that type of cluster is here: https://cs.uwaterloo.ca/twiki/view/UW/Paper

Contacts for questions/issues with the cluster are Harsh Roghelia (hkroghel@uwaterloo.ca) and Lori Paniak (ldpaniak@uwaterloo.ca) via email or UW MS Teams chat.

To start a Teams chat with Harsh Roghelia, click HERE

To start a Teams chat with Lori Paniak, click HERE

Principal Investigators

Tamer Ozsu and Khuzaima Daudjee

Cluster hardware overview

system count cpu memory disk interconnect other OS BIOS version Hyperthreading
tem01-tem04 4 2x Intel E5-2620v2 @2.1GHz (12 physical cores total) 32GB 2x 1TB 7200RPM HDD GbE + 10GbE   Ubuntu 16.04.3/4.10 kernel 3.2 Enabled
tem05-tem06 2 2x Intel E5-2620v2 @2.1GHz (12 physical cores total) 32GB 400GB Intel S3700 SSD + 1TB 7200RPM HDD GbE + 10GbE   Ubuntu 16.04.3/4.10 kernel 3.0 Disabled
tem07-tem82 76 2x Intel E5-2620v2 @2.1GHz (12 physical cores total) 32GB 2x 1TB 7200RPM HDD GbE + 10GbE   Ubuntu 16.04.3/4.10 kernel 3.0 Disabled
tem83-tem130 48 2x Intel E5-2620v2 @2.1GHz (12 physical cores total) 32GB 2x 1TB 7200RPM HDD GbE + 10GbE   Ubuntu 16.04.3/4.10 kernel 3.2 Enabled
tem201-tem204 4 4x Intel E5-4620 v2 @ 2.60GHz (32 physical cores total) 256GB 400GB Intel S3700 SSD + 2x 1TB 7200RPM HDD GbE + 10GbE + 40GbE   Ubuntu 16.04.3/4.10 kernel 3.0a Disabled
tem205-tem207 3 4x Intel E5-4620 v2 @ 2.60GHz (32 physical cores total) 256GB 400GB Intel S3700 SSD + 2x 1TB 7200RPM HDD GbE + 10GbE + 40GbE   Ubuntu 16.04.3/4.10 kernel 3.0b Enabled
tembo.cs 1 2x Intel E5-2630 v2 @ 2.60GHz (12 physical cores total) 128GB 2x 1TB 7200RPM HDD (OS) + 24x 3TB 7200RPM SAS HDD (ZFS) 2x 10GbE + 40GbE 10GbE UW uplink, ZFS storage array for user homes via 10GbE NFS, OpenVPN gateway Ubuntu 16.04.3/4.4 kernel 3.0a Enabled

Network configuration

system gigabit interface 10 gigabit interface 40 gigabit interface Mellanox firmware revision
tem[01-130] eno1: 192.168.152.0/24 enp4s0: 192.168.252.0/24 N/A 2.40.7000
tem20{1-7] enp193s0f0: 192.168.152.0/24 enp196s0: 192.168.252.0/24 enp2s0: 192.168.240.0/24 2.40.7000

Switch configuration

switch model OS/firmware version IP address
rack 1 gigabit Supermicro G2252 2.0.0.11 192.168.152.2
rack 2 gigabit Supermicro G2252 2.0.0.11 192.168.152.3
rack 3 gigabit Supermicro G2252 2.0.0.11 192.168.152.4
rack 1 10/40/56 gigabit Mellanox SX1710 3.6.4112 192.168.152.8
rack 2 10/40/56 gigabit Mellanox SX1024 3.6.4112 192.168.152.6
rack 3 10/40/56 gigabit Mellanox SX1024 3.6.4112 192.168.152.7
rack 1 40/56 gigabit Mellanox SX1012 3.6.4112 192.168.152.5

Password changes

The cluster uses a Samba4 Active Directory system to manage users and system access. Passwords can be changed when logged into the head node of the cluster by issuing the following command:

samba-tool user password

You will be prompted for your current password and then a new password (meeting complexity), twice:

Password for [LOCAL-DOMAIN\ldpaniak]:
New Password: 
Retype Password: 

Reservation System

https://tembo-reserve.cs.uwaterloo.ca/Web/?

The Tembo cluster features an autonomous, user-managed graphical interfaces for reserving Facility resources. The Reservation System allows for users to reserve multiple compute resources easily and include other Tembo users in reservations to facilitate collaboration. The Tembo Reservation System allows users to reserve resources for 30-minute periods on the top and bottom of the hour. Reservations can be placed, modified and deleted until 5 minutes before the opening of the next reservation period. Once a reservation has been successfully registered, all participants will receive a notification e-mail with details of the reservation. For long-term reservations spanning several days, it is highly-recommended to generate a series a repeating 1-day reservations and include your contact information in the reservation description.

At the end of the reservation period, access to reserved resources will be revoked and reassigned to users as allocated in the Reservation System.

Utilities

pdsh

Parallel distributed shell (pdsh) allows one to send commands to many machines in parallel. From tembo.cs you can use it with a file of hostnames eg.

pdsh -F hostfile -l user "command"

There are pre-defined host patterns in /etc/genders which also gives the format for the hostfile.

pdsh -g yellow "df -h"

man pdsh is useful.

Open VPN

All-in-one configuration file can be loaded by most OpenVPN clients: https://cs.uwaterloo.ca/twiki/pub/CF/Tembo/tembo.ovpn

Troubleshooting

If you notice that you are unable to login to a reserved machine or if there is any other issue with one of the machines, please email Harsh Roghelia (hkroghel@uwaterloo.ca) or Lori Paniak (ldpaniak@uwaterloo.ca), or you can start a Teams chat with either as well:

To start a Teams chat with Harsh Roghelia, click HERE

To start a Teams chat with Lori Paniak, click HERE

Topic attachments
I Attachment History Action Size Date Who Comment
Unknown file formatovpn tembo.ovpn r1 manage 2.0 K 2021-07-07 - 10:52 HarshRoghelia  
Edit | Attach | Watch | Print version | History: r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r1 - 2021-07-07 - HarshRoghelia
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback