# R630 DIMM B2 reseat procedure

**Last Updated:** 2026-01-31  
**Document Version:** 1.0  
**Status:** Active Documentation

---

**Use when:** You have an alert or message to reseat DIMM B2 on one of the R630s.

---

## 1. Identify which R630

| Proxmox host | IP address   | Role / notable workloads |
|--------------|-------------|---------------------------|
| **r630-01**  | 192.168.11.11 | NPMplus (10233), many infra containers (69 LXC) |
| **r630-02**  | 192.168.11.12 | besu-rpc-public-1 (2201), Blockscout (5000), Firefly, MIM (10 LXC) |

- If the alert shows **hostname**: `r630-01` = .11, `r630-02` = .12.
- If the alert shows **IP**: 192.168.11.11 = r630-01, 192.168.11.12 = r630-02.
- If the alert shows **Dell service tag / iDRAC**, match that to the physical server you have labeled as r630-01 or r630-02.

---

## 2. Impact of taking that host down

- **r630-01 down:** NPMplus (public proxy) and many services unreachable until host is back. Plan for a short maintenance window.
- **r630-02 down:** Public RPC (besu-rpc-public-1), Blockscout, Firefly, MIM API unreachable until host is back.

DIMM reseat requires **power off**. There is no live reseat for memory on R630.

---

## 3. Pre-maintenance (optional but recommended)

1. **Notify** anyone using services on that host.
2. **Optional:** Migrate or shut down VMs/containers you can move (Proxmox cluster) to reduce I/O before power off.
3. **Optional:** Put the node in maintenance in Proxmox (Datacenter → select node → Maintenance) so the cluster doesn’t try to start resources on it during the work.

---

## 4. Reseat DIMM B2 – steps

1. **Shut down the R630** (from Proxmox: shutdown the node, or from iDRAC: Power → Power Off).
2. **Power off at the PSU** (or pull power) and wait ~30 seconds. Touch a grounded chassis part to discharge static.
3. **Open the chassis** and locate memory. On Dell R630:
   - **B2** = channel **B**, slot **2** (see Dell R630 Owner’s Manual / memory population rules for exact slot position).
   - Slots are usually labeled on the board (e.g. A1–A4, B1–B4, etc.).
4. **Reseat B2:**  
   - Release the ejector clips at both ends of the DIMM.  
   - Remove the module, then reinstall it firmly until the clips click.  
   - Ensure the notch aligns and the module is fully seated.
5. **Close the chassis**, restore power, and **power on** the server.
6. **Verify:**  
   - Enter BIOS/iDRAC and check **System Memory** (or run memory test if available).  
   - Once the OS is up, from Proxmox or SSH: `ssh root@<host-ip> 'dmidecode -t memory | grep -A2 "Locator: B2"'` (or check total RAM with `free -h`) to confirm B2 is present and size is correct.

---

## 5. After maintenance

1. Exit **maintenance mode** on the node in Proxmox if you used it.
2. Confirm **pveproxy**, **pvedaemon**, **pvestatd** are active and **Web UI (8006)** is reachable.
3. Run a quick health check:  
   `./scripts/check-all-proxmox-hosts.sh`

---

## Quick reference

- **r630-01:** 192.168.11.11 — NPMplus, infra (69 LXC)  
- **r630-02:** 192.168.11.12 — RPC public, Blockscout, Firefly, MIM (10 LXC)  
- **DIMM B2:** Channel B, slot 2 — power off before reseat.  
- **Health check:** `./scripts/check-all-proxmox-hosts.sh`