--- title: "Redundant network for IOE Pulchowk campus" subject: "Internet and Intranet" chapter: 5 key: "semester_8/internet_and_intranet:redundant-network-ioe-pulchowk" source: https://www.bctengineer.com/q/semester_8/internet_and_intranet/redundant-network-ioe-pulchowk site: BCT Engineering --- # Redundant network for IOE Pulchowk campus Internet and Intranet · Chapter 5: Designing Internet Systems and Servers - Asked in: 2073 Magh, 12 marks - Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14 - Page: https://www.bctengineer.com/q/semester_8/internet_and_intranet/redundant-network-ioe-pulchowk - Whole chapter: https://www.bctengineer.com/md/semester_8/internet_and_intranet/chapter-5 ## Question Design the redundant network for IOE Pulchowk campus with 4 departments including ISP department which are around 400 m apart from each other. Three departments have 5 labs each with around 24 computers in each room. ISP contains server farm with server like DNS, DHCP, Email, FTP and webserver and the main internet router which will be connected to upstream provider. Propose appropriate equipment (L1, L2 and L3) and physical wires for the network design. ## Short answer Use single-mode OS2 fibre between buildings (3,000 m total) because 400 m exceeds copper limits. Each department gets two L3 distribution switches with routed links to a stacked core pair; access switches use LACP and RSTP. Deploy 55 boxes CAT6, 9 access switches, 6 distribution switches, 2 core switches, and 15 APs. ## Answer #### Assumptions - Four buildings: the ISP department (hub with server farm and main router) and three departments (Dept-1, Dept-2, Dept-3). Each department is about 400 m from the ISP building. This distance rules out copper (90 m limit), so the buildings are linked with fibre. - Each department building has 5 labs x 24 PCs = 120 PCs, plus 1 AP per lab (5 APs) = 125 drops. Average horizontal run 40 m. - Servers: DNS, DHCP, Email, FTP, Web. One upstream provider as stated, with a second link as an optional backup. #### Redundant design ```text Upstream ISP | [Edge Router] | [Firewall] | [Core-1]====LACP====[Core-2] ISP building | | | | | | | | +--Server farm(2 SW)--+ | | | Fibre route A (400 m) Route B (400 m, other path) | | +-----+------+------+---------+------+ [Dist-1a/1b] [Dist-2a/2b] [Dist-3a/3b] one pair per dept | | | 3 access SW 3 access SW 3 access SW (48-port) Labs 1-5 Labs 1-5 Labs 1-5 ``` - **Core:** two L3 switches in the ISP building, stacked/LACP, so a core failure is invisible to users. - **Distribution:** each department has two L3 distribution switches. Each one is uplinked over a **different fibre route** to a different core switch, and OSPF routes between them. Because these links are routed (L3), there is no L2 loop. - **Access:** within each department building, access switches connect to both distribution switches by LACP and run RSTP with BPDU guard. - **Servers:** NIC teaming to two server-farm switches. The router uses HSRP/VRRP and a backup link. #### Equipment and wires | Layer | Item | Qty (total) | |---|---|---| | L1 | CAT6 UTP, 305 m box (375 drops x 40 m x 1.1 = 16,500 m) | 55 boxes | | L1 | Single-mode OS2 6-core armoured fibre, 2 routes x 3 buildings x 500 m (400 m + slack) | 3,000 m | | L1 | Patch panels 24-port CAT6 (6 per building) | 18 | | L1 | Fibre patch panel 12-port LC | 8 | | L1 | CAT6 patch cords 2 m / 1 m | 375 / 375 | | L1 | LC-LC fibre patch cords, 10G SFP+ LR modules | 24 / 28 | | L1 | Faceplates and jacks (single port) | 375 | | L2 | 48-port PoE access switch (3 per building) | 9 | | L2 | 24-port server-farm switch | 2 | | L3 | Distribution L3 switch 24-port + SFP+ (2 per building) | 6 | | L3 | Core L3 switch (ISP building) | 2 | | L3 | Edge router, firewall | 1, 1 (+1 each for HA) | | Other | Wireless AP (5 per department) | 15 | | Other | Servers: DNS, DHCP, Email, FTP, Web | 5 | | Other | Racks, UPS (one pair per building) | 4, 4 | The fibre is single-mode so that 10 Gbps works over 400 m and later upgrades are easy; multimode OM3 supports 10G only up to about 300 m, so it is not enough. #### IP plan (10.10.0.0/16) Each lab has 24 PCs, a gateway and an AP, so a /27 (30 hosts) is used. | Block | Subnet | Notes | |---|---|---| | Dept-1 labs 1-5 | 10.10.1.0/27, .32, .64, .96, .128 (/27 each) | VLAN 11-15 | | Dept-1 Wi-Fi / mgmt | 10.10.1.160/27, 10.10.1.224/27 | | | Dept-2 labs 1-5 | 10.10.2.0 - 10.10.2.128 (/27 each) | VLAN 21-25 | | Dept-3 labs 1-5 | 10.10.3.0 - 10.10.3.128 (/27 each) | VLAN 31-35 | | Server farm DMZ (Web, Mail, FTP) | 10.10.10.0/27 | | | Internal servers (DNS, DHCP) | 10.10.10.32/27 | | | Core-distribution links | 10.10.250.0/30 each | OSPF area 0 | DHCP is relayed from the server farm to every lab VLAN. A single Internet-facing address block from the provider is used for NAT and the public servers. ## Check yourself The answer's key points as multiple-choice questions, with the correct option and why. 1. Why does the design link the departments to the ISP building with fibre instead of copper? A. Fibre is cheaper per metre than CAT6 B. The buildings are about 400 m apart, far beyond the 90 m copper limit C. CAT6 cannot carry Ethernet traffic at all D. Fibre removes the need for any switches Answer: B. UTP copper works only up to about 90 m for a horizontal run, so a 400 m link between buildings must use fibre. (Key point: Fibre between buildings) 2. How is redundancy provided between each department and the core? A. One distribution switch with a single long fibre B. Two distribution switches per department, each uplinked over a different fibre route to a different core switch C. Two access switches connected to each other only D. A wireless link replaces the fibre if it fails Answer: B. Each department has a pair of L3 distribution switches on separate fibre routes to separate cores, so no single cable or switch failure isolates it. (Key point: Redundant distribution uplinks) 3. Why is there no Layer 2 loop on the distribution-to-core links? A. Because the links are routed (L3) and OSPF is used between them B. Because all links are disabled except one C. Because hubs are used D. Because fibre cannot form loops Answer: A. The core-distribution links are routed, so OSPF handles path selection and no L2 loop forms. RSTP and BPDU guard are used only inside the building at the access layer. (Key point: Loop avoidance) 4. Why is single-mode OS2 fibre chosen rather than multimode OM3? A. OM3 is not available in Nepal B. OM3 reaches 10G only about 300 m, which is less than the 400 m needed C. Single-mode fibre carries only voice D. Multimode fibre needs no connectors Answer: B. Multimode OM3 supports 10G only to about 300 m, so it cannot cover 400 m; single-mode does and allows later upgrades. (Key point: Single-mode fibre choice) 5. Why is a /27 subnet used for each lab? A. It gives 30 usable hosts, enough for 24 PCs, a gateway and an AP B. It gives 254 hosts for future growth C. It gives only 14 hosts and saves addresses D. It is the only mask that works with VLANs Answer: A. A /27 has 30 usable addresses, which covers 24 PCs plus the gateway and AP. (Key point: Lab subnet sizing) ## Related in Internet and Intranet - [Redundant network design for insurance company](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Aredundant-network-xyz-insurance) — Internet and Intranet, chapter 5 - [BoQ for IOE ICT building six labs](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Aboq-ioe-ict-six-labs) — Internet and Intranet, chapter 5 - [Redundant hierarchical network design for bank](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Aredundant-network-bank-four-departments) — Internet and Intranet, chapter 5 - [BOQ for DOECE four computer labs](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Aboq-doece-four-labs) — Internet and Intranet, chapter 5 - [Design network for two buildings with fiber](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Anetwork-design-guidelines-two-buildings) — Internet and Intranet, chapter 5 - [BoQ for NITC four research labs](https://www.bctengineer.com/question.md?key=semester_8%2Finternet_and_intranet%3Aboq-nitc-research-labs) — Internet and Intranet, chapter 5 ## Related in other subjects - [Three transducers to control room: routing and S/H](https://www.bctengineer.com/question.md?key=semester_5%2Finstrumentation_ii%3Astepper-transducers-sh) — Instrumentation II, chapter 4 - [Subnetting for Banijya bank departments](https://www.bctengineer.com/question.md?key=semester_7%2Fcomputer_networks_and_security%3Asubnetting-banijya-bank) — Computer Networks and Security, chapter 4