user@netcalc ~/calc $
ipv4_calculator
Calculate network, broadcast, mask, hosts, class and binary for any IPv4 address
ip:
mask:
✓ stdout — ipv4_calculator
Enter runs the calculation · Shift+Enter inserts a new line
user@netcalc ~/calc $
ipv6_calculator
Calculate network info for IPv6 addresses in CIDR notation
ipv6:
pfx:
✓ stdout — ipv6_calculator
user@netcalc ~/calc $
vlsm_calculator
Variable Length Subnet Mask — assigns optimal blocks for each network segment
net:
pfx:
// Subnets required
name:
hosts:
name:
hosts:
name:
hosts:
✓ stdout — vlsm_calculator
| # | Name | Hosts Req | Hosts Avail | Prefix | Network | First Host | Last Host | Broadcast | Mask |
|---|
user@netcalc ~/calc $
usable_ip_range
Shows the exact usable host range for a given network
ip:
mask:
✓ stdout — usable_ip_range
user@netcalc ~/calc $
binary_ip_calc
Convert IP addresses to binary and vice versa, with per-octet visualization
dec:
bin:
✓ stdout — binary_ip_calc
user@netcalc ~/network $
cidr_converter
Convert between CIDR, decimal mask and wildcard. Includes full /1–/32 reference table.
in:
✓ stdout — cidr_converter
📋 CIDR Reference Table
| CIDR | Subnet Mask | Wildcard | Hosts | /24 subnets | Block |
|---|
user@netcalc ~/network $
wildcard_mask_calc
Calculate the wildcard (inverse) mask for Cisco ACLs, OSPF and firewall rules
mask:
✓ stdout — wildcard_mask_calc
user@netcalc ~/network $
supernet_calculator
Find the minimum supernet containing all specified networks (route summarization)
nets:
✓ stdout — supernet_calculator
user@netcalc ~/network $
subnet_splitter
Split a network into equal subnets by specifying a new prefix
net:
mask:
new:
✓ stdout — subnet_splitter
| # | Network | First Host | Last Host | Broadcast | Hosts | Mask |
|---|
user@netcalc ~/network $
range_to_cidr
Convert an IP range (start–end) to the minimum set of CIDR blocks
start:
end:
✓ stdout — range_to_cidr
| # | CIDR Block | Network | Broadcast | Hosts |
|---|
user@netcalc ~/network $
ip_to_integer
Convert IPv4/IPv6 addresses to integer and vice versa
ip:
int:
✓ stdout — ip_to_integer
user@netcalc ~/network $
ipv6_normalize
Expand and compress IPv6 addresses — normalize any valid format
ipv6:
✓ stdout — ipv6_normalize
user@netcalc ~/network $
hex_ip_converter
Convert between hexadecimal and IP address — useful for Windows route tables and Cisco logs
ip:
0x:
✓ stdout — hex_ip_converter
user@netcalc ~/network $
vlan_calculator
802.1Q VLAN info: tag encoding, PCP, DEI, VID, range and compatibility
vlan:
pcp:
dei:
✓ stdout — vlan_calculator
user@netcalc ~/dns $
reverse_dns_generator
Generate the PTR in-addr.arpa zone for an IPv4 network
net:
pfx:
dom:
✓ stdout — reverse_dns_generator
user@netcalc ~/dns $
ipv6_ptr_generator
Generate the PTR ip6.arpa record for an IPv6 address
ipv6:
host:
✓ stdout — ipv6_ptr_generator
user@netcalc ~/dns $
ip_geolocation
Locate a public IP or resolve a domain and geolocate its public address
Domains are resolved through DNS first; geolocation is based on the resulting public IP.
geo:
✓ stdout — ip_geolocation
user@netcalc ~/dns $
ip_address_planner
Interactive IP address space planner — click cells to edit, export to CSV
📋 IP Address Plan
| # | Name | Network | Mask/CIDR | Gateway | VLAN | Location | Notes |
|---|
user@netcalc ~/legal $
Privacy
Data stored in your browser
Theme, language, promotion dismissal and calculation history are stored in localStorage on this device.
Calculations
Calculations are sent through this site's same-origin Worker proxy to the private calculation service.
Performance metrics
LCP, CLS and INP may be stored as aggregate performance measurements by tool and language. Calculation inputs and persistent user identifiers are not included.
GeoIP
Only GeoIP lookups are processed by an external provider. The requested IP address is sent for that lookup.
Issue reports via Telegram
A report link opens the public Telegram bot and includes only the current page slug in the bot start parameter. Calculator inputs and report text are not sent by this website. Text you choose to send inside Telegram is handled by Telegram and the bot.
Cookies and advertising
Advertising is controlled by the production configuration. When Google AdSense is enabled, consent is handled through a certified CMP before personalized advertising is used where required.
Delete local data
Use the button below to clear calculation history. To remove every local preference, clear this site's data in your browser settings.
This page provides operational information; it is not a legal-compliance certification.
Asymmetric routing and stateful firewalls: why valid traffic becomes INVALID
Understand how different forward and return paths break state tracking and how to prove the asymmetry before weakening firewall rules.
01
Stateful firewalls expect a coherent conversation path
Connection tracking learns a flow from packets it sees. If the forward path crosses one router but the return path bypasses it, the first firewall may see only half the conversation and classify later packets as INVALID or out of state.
02
Prove asymmetry before weakening INVALID drops
Temporarily logging state, interfaces and five-tuples is safer than simply accepting INVALID traffic. Compare traceroutes, routing tables and packet captures on both directions until you can name the device that sees only one side of the flow.
03
ECMP and backup paths can make the symptom intermittent
Equal-cost routing, policy routing and failover links can move only some flows to a different return path. That creates failures that appear random even though the routing decision is deterministic for each hash or policy condition.
04
Fix routing symmetry or state placement first
The cleanest solution is usually to make both directions cross the same stateful boundary, or place stateful filtering where both directions are guaranteed to pass. Broad exceptions for INVALID traffic should be the last resort and narrowly scoped.
05
Use packet capture timestamps to pair both directions
Capturing on the suspected firewall and on an adjacent router can show that SYN leaves through one device while SYN/ACK returns somewhere else. Matching five-tuples and timestamps is stronger evidence than a single traceroute.
06
Policy routing can hide the asymmetry in only one direction
Source-based rules, VRFs and marking can override the main routing table. Inspect the policy decision that applies to the exact source and connection mark rather than relying only on a generic route lookup.
07
Keep security controls narrow during diagnosis
If an exception is unavoidable for testing, restrict it by interface, source and destination and remove it after proving the path. Turning a global INVALID drop into accept can hide unrelated spoofing or stale-state problems.
EX
Worked example
- Capture a new TCP connection at the stateful firewall.
- Record ingress/egress interface for SYN and look for the returning SYN/ACK.
- Run route lookups for both source→destination and destination→source with the real policy context.
- Correct routing so both directions cross the same state boundary, then retest with the normal INVALID drop enabled.
✓
Operational checklist
- Prove where each direction travels.
- Check policy routing and ECMP, not only the main table.
- Keep diagnostic firewall exceptions narrowly scoped.
- Restore normal INVALID handling after the route is fixed.
Tools to apply it
Standards and references
🕐 History