This is an old revision of the document!
Table of Contents
Infrastructure Hardware
This page covers all other hardware that doesn't fit under the umbrella of testnode hardware.
gw1/gw2
Summary
These two hosts are 1U custom-built Supermicro systems.
gw1 was repurposed as store03 to use as an arbiter node for the Gluster storage backing RHEV VMs.
gw2 is not powered on or accessible
Hardware Specs
| Count | Manufacturer | Model | Capacity | Notes | |
|---|---|---|---|---|---|
| Chassis | N/A | Supermicro | N/A | ||
| Mainboard | N/A | Supermicro | PDSMi+ | N/A | |
| CPU | 1 | Intel | Xeon X3220 @2.40GHz | N/A | ARK |
| RAM | 4 DIMMs | ? | ? | 1GB | DDR2 667MHz 4GB total |
| HDD | 2 | HGST | HDS721010CLA330 | 1TB | Software RAID1 |
| SSD | |||||
| NIC | 2 | on-board |
store{01,02}.front.sepia.ceph.com
Summary
Serve as the main gluster storage nodes for the rhev cluster. They're running CentOS 7.
store01 was formerly gw.sepia.ceph.com and still hosts a number of services:
store02 was formerly mira123.
Hardware Information
Mira-type chassis with 16GB RAM and 1x Intel(R) Xeon(R) CPU X3430 @ 2.40GHz
Storage
- Populated with 7x 4TB Hitachi HDD
- 20TB Areca RAID5
- One bay is not used in each system so if desired, the stripe size and RAID level could be changed
Networking
- store01 has a 1Gb bridge on the ipmi network from when it was the OpenVPN gateway.
- store01 has a 10Gb uplink with address 8.43.84.133 mainly as a backup “in” to the lab. This is leftover from when the host was the OpenVPN gateway.
- Both nodes have an add-on dual-port 10Gb SFP+ card with both ports cabled and configured as bond0 (bond type 6)
store03.front.sepia.ceph.com
Summary
Sole purpose is to serve as an arbiter node for the Gluster volume backing rhev VMs.
store03 was formerly gw1.
Hardware Information
ssdstore{01,02}.front.sepia.ceph.com
Summary
Serve as new NVMe Gluster storage nodes for the rhev cluster. They're running RHEL 7. Purchased/donated by OSAS.
Hardware Information
| Count | Manufacturer | Model | Capacity | Notes | |
|---|---|---|---|---|---|
| Chassis | N/A | Supermicro | N/A | ||
| Mainboard | N/A | Supermicro | PDSMi+ | N/A | |
| CPU | 1 | Intel | Xeon X3220 @2.40GHz | N/A | ARK |
| RAM | 4 DIMMs | ? | ? | 1GB | DDR2 667MHz 4GB total |
| HDD | 2 | HGST | HDS721010CLA330 | 1TB | Software RAID1 |
| SSD | |||||
| NIC | 2 | on-board |
teuthology.front.sepia.ceph.com
Hosted previous teuthology service for Sepia lab. Running Trusty with a custom kernel so the LRC can be mounted. Now only up for legacy purposes. Can be moved to another rack and/or repurposed.
Mira-type chassis with 32GB RAM and 2x Intel(R) Xeon(R) X3440 @ 2.53GHz
Storage
- Populated with 8x 4TB Hitachi HDD
- 80GB Areca RAID6 hosts
/ - 24TB Areca RAID6 hosts
/home
Networking
- 1x cabled but unused 1Gbps link (eth0)
- 1x 10Gbps uplink for all traffic (eth3)
pulpito.front.sepia.ceph.com
Hosts pulpito and paddles. Running Ubuntu Precise.
Mira-type chassis with 16GB RAM and Intel(R) Xeon(R) X3440 @ 2.53GHz
Storage
root@pulpito:~# cli64 vsf info # Name Raid Name Level Capacity Ch/Id/Lun State =============================================================================== 1 ARC-1222-VOL#000 Raid Set # 000 Raid6 400.0GB 00/00/00 Normal 2 ARC-1222-VOL#001 Raid Set # 000 Raid1+0 60.0GB 00/00/01 Normal 3 ARC-1222-VOL#002 Raid Set # 000 Raid6 5510.0GB 00/00/02 Normal =============================================================================== root@pulpito:~# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT sda 8:0 0 372.5G 0 disk ├─sda1 8:1 0 356.5G 0 part / ├─sda2 8:2 0 1K 0 part └─sda5 8:5 0 16G 0 part [SWAP] sdb 8:16 0 55.9G 0 disk /var/lib/postgresql sdc 8:32 0 5T 0 disk
Networking
- 1x cabled 1Gbps link (eth0)
gitbuilder.ceph.com
Summary
AKA apt-mirror or gitbuilder-archive.
Services
- Serves packages built by gitbuilders
Storage
- 8x 4TB HGST drives
- Volume set 1 is a 80GB RAID6 for the OS
- Volume set 2 is a 24TB RAID6 for /home where all the data for services listed above live
Networking
- 1x 10Gb uplink at address 8.43.84.130
Temporary VPN tunnel for drop.ceph.com access to the longrunningcluster(Must be restarted1)service openvpn restartanytime the VPN service is restarted)
hv{01..03}
Summary
Main RHEV highly available hypervisor nodes. Running RHEL7.
Infra admins have standard ssh access as their user.
Hardware Specs
| Count | Manufacturer | Model | Capacity | Notes | |
|---|---|---|---|---|---|
| Chassis | N/A | Supermicro | SYS-1028U-TRT+ | N/A | |
| Mainboard | N/A | Supermicro | X10DRU-i+ | N/A | |
| CPU | 2 | Intel | E5-2660 v3 @ 2.60GHz | N/A | ARK |
| RAM | 8 DIMMs | Hynix Semiconductor | HMA42GR7MFR4N-TF | 16GB | 128GB total |
| HDD | 0 | ||||
| SSD | 4 | Intel | SSDSC2BB480G6 | 480GB | Software RAID5 |
| NIC | 2 ports | Intel | X540-AT2 | 10Gbps | On-board RJ45. Only useful for 1Gb due to no 10GBASE-T switches |
| NIC | 2 ports | Intel | 82575EB | 1Gbps | Add-On NIC. Not used |
| NIC | 2 ports | Intel | 82599ES | 10Gbps | SFP+ Port 1 for LAN / Port 2 for WAN |
| NIC | 2 ports | Intel | 82599ES | 10Gbps | SFP+ Port 1 for rhevm network / Port 2 unused |
Dual PSU redundantly cabled to two PDUs.
