Skip to content
VCloud
Menu
INTERNET · WHOLESALE TRANSIT

Full-table BGP transit, on a network that filters its routes.

Wholesale IP transit on AS137492 — multi-homed across two Tier-1 upstreams and PKIX, with RPKI ROA validation, MANRS-aligned filtering and a community set that lets you traffic-engineer without opening a ticket.

AS137492 IPv4 + IPv6 dual-stack RPKI + MANRS
VCloud Upstream A Upstream B IXP / PKIX Hyperscaler AS137492 AS3257* AS6939* AS24482* AS15169* eBGP MD5 eBGP RPKI Multi-LAG PNI 100G CUSTOMER EDGE · IPv4 + IPv6 · BGP4 · 95TH PERCENTILE BILLING * UPSTREAM AS NUMBERS SHOWN ARE ILLUSTRATIVE
OVERVIEW

Transit, sold the way an engineer would buy it.

IP transit is for customers who run their own AS and want a clean BGP feed: a full v4 and v6 table, MD5 or RFC 8950 next-hop, MED and local-pref handling that doesn't surprise you, and a NOC that can read a route trace without rebooting the conversation. We are multi-homed across two Tier-1 upstreams and present at PKIX, with private peering into the major content networks active in Pakistan.

Sessions are eBGP over your handoff port, MD5-authenticated by default. You announce your prefixes — they must be either authenticated by RPKI ROAs or registered against IRR objects we can validate — and we propagate them to all upstreams and peers. BFD is supported for sub-second failure detection on 10G and 100G ports. Communities cover prepending, scope (no-export to peers / customers / upstreams), blackhole, and graceful shutdown.

USE CASES

Customers we deliver transit to.

01

Regional ISPs and WISPs

Operators with their own AS who want a second or third upstream for resilience and traffic blending.

02

DC operators

Carrier-neutral data centres reselling transit to their tenants and needing a BGP session that doesn't go down on planned maintenance.

03

Content owners and OTT platforms

Local content networks pushing high egress to Punjab eyeballs who want predictable peering and selective announcement scope.

04

Cloud and hosting providers

Local IaaS/PaaS operators advertising customer prefixes who need a transit that won't drop their RPKI-invalid neighbour's announcements.

05

Banks and large enterprises with PI space

Organisations holding their own /22 or /23 of provider-independent IPv4 who want a clean dual-stack BGP feed.

06

Network engineering teams in transition

Customers migrating off a poor incumbent who need a transit they can lab in parallel and cut over with a published session-config sample.

SPECIFICATION

Session and policy details.

BGP session
Metric Target Notes
Our AS AS137492 PeeringDB AS137492
Address families IPv4 + IPv6 Single or split sessions
Authentication MD5 (default) TCP-AO on request
Max prefixes 1M v4 / 250k v6 Per peer; tuned per customer
BFD Supported, 50/150 ms On 10G+ ports
Looking glass lg.vcloud.com.pk Public, read-only
Routing policy
Metric Target Notes
RPKI Drop invalids Per RFC 6811
IRR filtering RADb / RIPE / APNIC AS-SET required for >5 prefixes
MANRS actions All four Anti-spoofing, filtering, coordination, GVA
Communities Prepend / no-export / blackhole Documented in welcome email
DDoS blackhole 65500:666 → /32 + /128 Propagated to upstreams
Graceful shutdown GRACEFUL_SHUTDOWN (65535:0) Both directions
DEPLOYMENT

From quote to live session.

  1. 01

    Pre-flight

    Day 0 – 3. ASN, prefixes, IRR/ROA check, max-prefix and policy review.

  2. 02

    Port assignment

    Day 3 – 7. Cross-connect or LR optic delivery if you're in our DC, or PE port if you're on-net via fibre.

  3. 03

    Session up

    Day 7 – 10. eBGP brought up in test mode, prefixes audited, communities tested.

  4. 04

    Production

    Day 10 – 14. Customer announces, traffic ramps, monthly 95th billing starts.

FAQ

What network engineers ask.

What's your filtering posture?
We drop RPKI invalids in both directions. For prefix learning we require either a valid ROA or a current IRR object (RADb, RIPE, APNIC) matching your AS-SET. We won't accept un-registered space. For announcements out, we strip private ASNs, scrub bogons, and re-attach our community set.
How is billing structured?
95th percentile, calculated independently on inbound and outbound, billed on the higher of the two. Burstable up to the port speed; commit can be raised mid-term. Monthly statements include the bin chart and the calculated 95th.
Do you support flowspec?
Customer-driven flowspec to our edge is supported on a case-by-case basis after a security review — typically for content networks during attack mitigation. We do not propagate flowspec rules to our upstreams.
What's the smallest commit you'll quote?
1 Gbps commit on a 10G port is the floor for new transit customers. Below that, the pricing competes poorly with our B2B internet product, which we'd recommend instead.
Can I peer privately with you (instead of buying transit)?
Yes — open peering policy. Settlement-free PNI is available at our DCs in Lahore and Islamabad for networks with sufficient mutual traffic; otherwise we suggest meeting at PKIX.
What's your response time on a routing incident?
Our NOC is staffed 24×7. P1 routing incidents (BGP session down, RPKI fail, suspected hijack) are picked up inside 15 minutes and escalated to a senior engineer immediately. Mean time to first response on tickets last quarter was 6 minutes.
Is IPv6 a default or an option?
Default. Every customer port comes up dual-stack and we encourage IPv6 announcements. Inbound IPv6-only sessions are welcome.

Send us your AS-SET, prefixes and target commit.

We'll respond with a port quote, the welcome email content for your records, and an estimate of next-hop latency from the closest of our PoPs.