Practical knowledge. Your next chapter online.
JournalThe UK edition
Web hosting basics

100 Web Hosting Terms Explained for UK Website Owners

Hosting vocabulary becomes useful when it explains a real decision. This glossary covers 100 terms across infrastructure, domains, email, security and maintenance, correcting common misunderstandings about cloud hosting, SSL, backups and bandwidth so you can read a provider’s specification with confidence.

telaHosting editorial teamPublished 31 October 2026 · 21 min read
In this article 2 sections

Hosting documentation mixes networking, software, security and commercial language. This glossary explains 100 useful terms in plain English, with the distinctions that matter when choosing or operating a UK business website. It is a reference to consult alongside your provider’s actual service specification.

Some words have several meanings. Bandwidth is often used loosely in marketing, managed has no universal scope, and cloud does not automatically mean a website runs in several regions. Where a term describes a capability, confirm whether your chosen product actually supplies it.

100 terms, from Apache to zlib

1. Apache HTTP Server

Open-source web-server software that receives HTTP requests and serves responses, either directly from files or through applications. Its behaviour depends on configuration and modules. The software name alone does not determine the speed or security of a complete website.

2. API

An application programming interface defines how software components request information or actions from one another. A shop might use a delivery API to obtain shipping options. Authentication, rate limits and error handling are important when the website depends on an external service.

3. Autoscaling

Automatic adjustment of computing capacity according to configured rules or demand signals. It can add instances or change other resources when supported. The application must be designed for the process, and spending controls matter; autoscaling is not included in every cloud product.

4. Log files: access and error logs

Records produced by servers and applications. Access logs describe handled requests, while error logs record problems or diagnostic messages. Logging coverage varies. Protect logs because they can contain identifiers or sensitive operational information, and set appropriate retention and access controls.

5. Active Directory

Microsoft technology for managing identities and resources in an organisation’s network. It can support centralised account and permission administration. It is not a normal requirement for a small public website, although larger organisations may integrate website authentication with their identity systems.

6. Bandwidth

Technically, network capacity expressed as a rate, such as megabits per second. Hosting marketing sometimes uses the word for a monthly data-transfer allowance instead. Ask which meaning applies so you do not confuse the amount transferred with the rate of transfer.

7. Burstable bandwidth

A service arrangement permitting temporary network use above a normal allocation, subject to defined conditions. The available peak, duration and charging rules depend on the provider. It does not guarantee that the application can handle every traffic spike without another bottleneck.

8. Backup

A retained copy intended to support recovery after deletion, corruption or another incident. Useful backups cover the required files, databases and configuration. Their value depends on separation, retention and tested restoration, rather than simply having a second copy somewhere on the same system.

9. Bot

Software that performs automated online tasks. Search crawlers and uptime monitors can be useful, while other bots attempt abuse or consume resources. The practical task is to distinguish legitimate automation from harmful activity rather than assuming every automated request should be blocked.

10. BGP

Border Gateway Protocol exchanges routing information between networks. Routing decisions reflect policies and available paths; they are not simply a calculation of the fastest route for each website visitor. It is normally managed by network operators rather than an everyday hosting customer.

11. Byte

A unit commonly consisting of eight bits. File sizes and storage allowances are generally expressed in bytes and their multiples, while network rates often use bits per second. Decimal GB and binary GiB also differ, so read capacity measurements carefully.

12. Cache

A stored result or resource reused to avoid repeated work. Caches exist in browsers, applications, proxies and delivery networks. Correct expiry and invalidation are essential: a cached public image is useful, but serving one customer’s private account page to another is dangerous.

13. cPanel

A commercial hosting control panel offering interfaces for tasks such as files, databases, domains and email, depending on configuration. Its inclusion and available functions vary by service. It is an administration tool, rather than a guarantee about the resources or maintenance supplied.

14. Cloud hosting

Hosting supplied through cloud infrastructure or a managed service built on it. The label covers several operating models. Automatic scaling, multi-region resilience and usage-based billing are separate product or architecture choices, not universal properties of every service advertised as cloud hosting.

15. CDN

A content delivery network serves suitable content through distributed infrastructure, often caching public assets closer to users. Cache rules, geographic coverage and origin behaviour matter. A CDN can reduce repeated delivery work without automatically fixing slow database queries or every personalised page.

16. Cron job

A task scheduled using cron or a compatible scheduling interface. Websites may use scheduled tasks for imports, reports or housekeeping. Confirm the timing, permissions and failure reporting, and avoid overlapping resource-heavy jobs that interfere with the live application.

17. Database

An organised store of data accessed through database software. A CMS may keep articles, settings and user accounts there. Databases require access controls, suitable backups and efficient queries; copying website files alone is usually insufficient to recover a database-driven site.

18. Dedicated hosting

A service allocating a physical server to one customer. That customer can run one or several applications. Exclusive hardware provides control, but administration, backups and resilience still require planning. A single dedicated server can remain a single point of failure.

19. Domain name

A human-readable name resolved through DNS to information about services. Registration gives rights to use the name under the applicable rules and term. It does not automatically include a website, email hosting or a trademark, and renewal remains important.

20. Disk space

The storage capacity allocated to a service or account. Files, databases, logs and email may count differently according to the plan. Storage size alone does not describe processing capacity or application speed; check what is included and how usage is measured.

21. DNS propagation

An informal name for the period during which DNS changes become visible as caches expire and authoritative information is updated. DNS is not one global database pushed everywhere at once. Record lifetimes and resolver behaviour affect what different users observe.

22. Dynamic content

Content generated or selected using application logic, data or user context. It includes product searches, account pages and personalised responses. Dynamic content can still be cached where appropriate; the label does not inherently mean that a website must be slow.

23. Email hosting

A service for sending, receiving and storing email, often using your own domain. It can be separate from website hosting. Compare mailbox capacities, access methods, sending limits and operational features rather than assuming all included email addresses provide equivalent capabilities.

24. Encryption

A cryptographic process that protects information using keys. Data may be encrypted in transit or at rest, with different controls and threats involved. Encryption does not repair weak application permissions or protect information after an authorised endpoint has exposed it.

25. EPP or transfer authorisation code

A credential used in transfer processes for many generic domain extensions. EPP itself is a provisioning protocol, not merely the code. Procedures vary: current .uk registrar changes use the relevant TAG workflow, with announced standardisation changes due in February 2027.

26. Egress traffic

Data leaving a service or network, such as files sent to visitors or another platform. Some infrastructure products charge separately for it. When estimating costs, distinguish incoming data, outgoing data and transfers between regions or services under the provider’s actual rules.

27. Endpoint

A destination or interface at which communication occurs. In API documentation it often means a particular service URL; in security discussions it may mean a device such as a laptop. Read the context rather than assuming the term always describes hardware.

28. Firewall

A control that permits or blocks traffic according to rules. Network firewalls and web application firewalls operate at different levels. They can contribute to protection, but do not replace updates, secure configuration or checks on what an authenticated application user can do.

29. FTP

File Transfer Protocol moves files between systems. Traditional FTP does not protect credentials and data through encryption. Prefer an appropriate secure method supported by your service, such as SFTP or FTPS; SFTP and FTPS are different protocols despite their similar names.

30. Free hosting

Hosting offered without a direct hosting charge under a particular service model. Limits, advertising, support and permitted use vary. It can suit experiments, but an organisation should evaluate continuity, account ownership and export options before relying on it for important work.

31. Fallback server

A secondary server available when the normal service cannot be used. Its usefulness depends on current data, routing and tested activation. Simply owning a spare server does not guarantee seamless recovery or that the latest transactions will be present.

32. Failover

Switching service to an alternative component when the primary component fails or becomes unavailable. It can be automatic or manual. Detection, state consistency and recovery procedures determine the result; failover should be tested rather than described as a guarantee of uninterrupted service.

33. Gateway

A component linking systems or translating between communication environments. A network gateway and a payment gateway perform different jobs. In a website project, identify the specific service, its dependencies and who supports it instead of treating gateway as one standard hosting feature.

34. Geo-hosting

An informal term for locating hosting in a chosen geographical area. Location can affect latency and data-handling arrangements, but nationality of the supplier does not establish server location. Confirm the actual service and measure relevant performance rather than assuming proximity solves everything.

35. Git

A distributed version-control system used to record and collaborate on changes, commonly in source code. It helps track history and coordinate development. A repository is not automatically a complete website backup: databases, uploaded files and secrets need separate handling.

36. Apache Guacamole

A browser-accessible remote desktop gateway supporting protocols including RDP, VNC and SSH. It is sometimes shortened to Guacamole; it should not be described as a standard graphical password system called GUAC. Its deployment and access controls require appropriate administration.

37. Green hosting

A marketing description for hosting associated with environmental commitments. Evaluate the specific claim, evidence, scope and reporting period. Renewable electricity, efficiency measures and offsets describe different things; none should be assumed merely because a provider uses a green label.

38. Gzip compression

A compression format commonly used to reduce the size of suitable text responses in transit. Server and browser support determine its use. Already compressed images may gain little, and compression is only one part of improving page delivery.

39. Hosting plan

The service package defining allocations, supported features and commercial arrangements. Names such as Starter or Business are provider-specific. Compare the actual resources, responsibilities and limits rather than assuming similarly named plans from different suppliers represent equivalent services.

40. HTTP and HTTPS

HTTP is the request-and-response protocol underlying much of the web. HTTPS carries HTTP over a TLS-protected connection. It helps protect data in transit and authenticate the endpoint according to its certificate, but does not certify the business or application as trustworthy.

41. Hypervisor

Software that creates or operates virtual machines on underlying hardware. It helps separate virtual environments and allocate resources. The wider platform’s configuration, resource commitments and maintenance still influence reliability and performance; virtual separation is not an unlimited-capacity promise.

Controls intended to limit other sites embedding files served from your account. Implementations commonly inspect request information and have limitations. Use suitable access controls for private files; preventing casual image embedding is not equivalent to preventing all copying or unauthorised access.

43. Hybrid hosting

An architecture combining different infrastructure environments, such as cloud services and dedicated systems. The term is broad. Document data flows, identity, networking and operational ownership, because combining platforms adds interfaces that must be maintained as well as potential flexibility.

44. High availability

A design goal and set of measures intended to keep a service usable despite certain failures. Redundant components, health checks and tested recovery may contribute. Specify the failure scenarios and measurement period rather than using the phrase to mean that downtime is impossible.

45. IP address

An address used by Internet Protocol to deliver network traffic. IPv4 and IPv6 have different formats, and public and private addressing serve different roles. An address can change or be shared; it is not a permanent identity for a person.

46. IMAP

Internet Message Access Protocol lets email clients access and manage messages on a mail server. Clients may also cache or download copies. Synchronisation makes server-side changes important, so deletion and backup arrangements deserve attention when several devices use the same mailbox.

47. Indexing

In search, processing information about a resource for possible inclusion in search results. Crawling and indexing are related but different. A page being accessible or listed in a sitemap does not guarantee indexing, and indexing does not guarantee a particular ranking.

48. Inode

A filesystem structure holding information about a file or directory. Some hosting services limit file or inode counts as well as storage capacity. Many tiny cache files can therefore exhaust an allowance even when the account has substantial unused disk space.

49. Instance

A particular running virtual machine, service or application copy, depending on context. In cloud infrastructure it often refers to a virtual server. Record its configuration and dependencies so that recreating the instance does not depend on undocumented manual changes.

50. IPsec

A collection of protocols for protecting IP communications through mechanisms including authentication and encryption. It is often encountered in VPN arrangements. Correct configuration and key management matter, and it should not be confused with the HTTPS configuration of a public website.

51. LiteSpeed

A family of web-server products and associated technologies. A hosting service may use a particular edition and caching integration. Evaluate the complete configuration and measured workload rather than claiming that a software logo proves one site will outperform every alternative.

52. Load balancer

A component distributing requests among multiple backends according to configured rules. It can support capacity and availability, but backends must share necessary state appropriately. Health checks, session handling and the load balancer’s own resilience are part of the design.

53. Latency and network latency

Delay in an operation or communication path. Network latency concerns transmission and response delays, often measured in milliseconds. Full page loading also includes server processing and browser work, so a low ping time is not the same as a fast website.

54. Load testing

A controlled test of system behaviour under a defined workload. It helps identify limits when the traffic model, environment and measurements are meaningful. Obtain permission for the infrastructure involved, avoid harming live users and report actual conditions instead of invented scores.

55. Linux hosting

Hosting using a Linux-based operating system. It can run many web stacks, subject to the service configuration. The operating-system label does not establish which application versions, permissions or administration tools are available, so check compatibility with the software you intend to use.

56. Malware

Malicious software intended to compromise, disrupt or misuse systems and information. Website malware can appear in application code, extensions or uploaded files. Removal should include investigating access, patching the cause and verifying recovery, rather than merely deleting one visible symptom.

57. MySQL

A relational database management system commonly used in web applications. It organises data in structures accessed through queries. Application compatibility, supported versions, access permissions and backup methods matter when moving a database between hosting environments.

58. Migration

Moving a website or service between environments. Files, databases, configuration, scheduled tasks, DNS and email can each require attention. Testing, a change plan and a recovery route reduce avoidable disruption; migration does not automatically repair existing application faults.

59. Mail server

Software and infrastructure handling email functions such as receiving, relaying or storing messages. Different components may perform different roles. Delivery also depends on DNS, authentication and recipient systems, so a functioning website does not prove business email is configured correctly.

60. Middleware

Software between application components that supports communication or common functions. The exact meaning depends on the stack: it might describe request processing, integration or messaging. Identify the component and its maintenance requirements rather than treating middleware as one standard product.

61. Managed hosting

Hosting with administration tasks supplied by the provider under a defined scope. That scope can range widely. Ask who updates the operating system, control panel and application, who monitors incidents and who restores data; managed does not automatically mean every technical task is included.

62. Nginx

Web-server software commonly used to serve content, proxy requests and support load-balancing or caching configurations. Its role depends on the deployment. The presence of Nginx is a technical detail, not independent proof of a website’s speed or resilience.

63. Nameserver

A server providing DNS information in a particular role. Authoritative nameservers hold answers for a zone, while recursive resolvers obtain and cache answers for clients. Changing a domain’s authoritative nameservers can affect website, email and other records together.

64. Network security

The controls and practices used to protect networks and their traffic. These include access restrictions, segmentation, monitoring and maintenance. Protection is layered and shared across people and systems; a firewall feature alone does not establish a secure operating environment.

65. NIC

A network interface card, or more broadly a network interface, connects a system to a network. Interfaces may be physical or virtual. Their stated rate is one component of capacity and does not guarantee the end-to-end throughput a website visitor will receive.

66. Node

A participating device or system in a network or distributed application. A cluster node may have a defined operational role. It is a general term and should not automatically be read as Node.js, which is a particular JavaScript runtime.

67. PHP

A programming language widely used for server-side web applications, including many CMS installations. The server executes PHP code and returns an appropriate response. Use supported versions compatible with the application and test upgrades rather than assuming every old extension will work unchanged.

68. Plesk

A commercial platform for administering hosting and server services. Available tools depend on the edition, extensions and configuration. As with other control panels, licensing, application support and ongoing server management are separate questions to resolve in a hosting proposal.

69. Proxy and reverse proxy

A proxy intermediates communication. A forward proxy acts on behalf of clients; a reverse proxy fronts one or more servers. Reverse proxies can handle routing, TLS or caching. The effect depends on configuration, and private data must be handled carefully.

70. Patch management

The process of identifying, testing and applying software updates. It includes knowing which components exist and how urgent fixes are handled. Balance safe testing with timely security response, retain recovery options and avoid leaving unsupported components in production indefinitely.

71. Ping

A diagnostic tool commonly using ICMP to test reachability and round-trip time. Some systems restrict its traffic, so a failed ping does not necessarily mean a website is down. A successful ping also does not prove the application is responding correctly.

72. Reseller hosting

A service enabling a customer to create or supply hosting accounts to others. The reseller normally manages its customer relationship while the upstream provider supplies specified infrastructure. Account limits, support boundaries, billing and responsibility for abuse need clear arrangements.

73. Redundancy

Additional components or paths intended to reduce dependence on one element. Redundant storage, power and servers address different failures. They help only when the wider design and procedures use them effectively, and they do not replace independent recovery copies.

74. Root access

Highly privileged administration access on Unix-like systems. It permits changes that ordinary accounts cannot make and can also cause extensive damage. Use restricted everyday accounts and controlled elevation where practical, with strong authentication and a record of administrative responsibilities.

75. RAID

A family of arrangements combining storage devices for different performance or fault-tolerance goals. Not every RAID level provides redundancy. RAID is not a backup: accidental deletion, corruption or a wider incident can affect the live data across the array.

76. Replication

Maintaining copies of data across systems, synchronously or asynchronously. It can support availability or other operational goals. Errors and unwanted changes may also replicate, so retention, consistency and recovery requirements still need a separate plan.

77. Scalability

The ability to accommodate changed demand by adjusting resources or architecture. Vertical scaling changes a system’s capacity; horizontal scaling adds systems. Applications, databases and operational processes must support the approach, and growth may introduce cost or complexity rather than being automatic.

78. Shared hosting

Hosting where customer accounts use a common server platform under defined limits and controls. It often reduces administration effort for standard websites. Resource allocation and isolation differ by provider, so shared does not mean every account receives an identical unrestricted portion.

79. SSL certificate or TLS certificate

A digital certificate used in establishing an authenticated TLS connection. SSL remains common marketing language, although modern HTTPS uses TLS. Check hostname coverage and validity. A certificate does not demonstrate that all website content, software or business practices are safe.

80. Snapshot

A point-in-time representation of a system, volume or other resource. Its coverage and consistency depend on the technology and procedure. It may depend on underlying storage and may not replace an independent backup, especially for recovery from a platform-wide incident.

81. Spam filter

A system evaluating messages to reduce unwanted or harmful email. Filters can miss abuse or incorrectly flag legitimate messages. Users still need safe handling practices, and administrators should review delivery problems rather than assuming a filter guarantees an entirely safe inbox.

82. Static content

Content served without being generated anew by server-side application logic for each request. Static files can still support interactive browser features through JavaScript. Static delivery can simplify hosting, but updates, forms and external services still require deliberate choices.

83. Traffic

A context-dependent term for activity or data moving through a website or network. Analytics visits, HTTP requests and transferred bytes are different measures. State which one you mean when reporting growth or investigating resource use, rather than treating them as interchangeable.

84. TTL

Time to live. For a DNS record, it indicates how long a cached answer may normally be reused. Lowering it before a planned change can help, but existing caches retain previously supplied lifetimes; it is not an instant global refresh button.

85. Two-factor authentication

Authentication using two distinct factor types, such as a password and a security key. Two passwords are not two different factors. Enable suitable protection on important accounts and retain recovery arrangements so stronger login controls do not leave the business dependent on one person.

86. Traceroute

A diagnostic tool exploring the network path toward a destination using responses from intermediate systems. Results can be incomplete or affected by filtering and routing. An unresponsive hop is not by itself proof that the hop is causing a website problem.

87. Throughput

The observed rate of useful work or data transfer over time. It differs from theoretical network capacity and from the total amount transferred in a billing period. Application processing, network conditions and protocol overhead can all limit the result.

88. Tokenisation

Replacing sensitive information with a substitute token handled under a particular system. Payment tokenisation can reduce the exposure of card details in a website’s workflow. The protection depends on the implementation and does not automatically remove all security or compliance responsibilities.

89. Uptime

The proportion of a defined period during which a service meets an availability criterion. The criterion and measurement point matter. A responsive homepage does not necessarily mean checkout works, and an advertised percentage is not a substitute for monitoring relevant business functions.

90. URL

A uniform resource locator identifies a resource through components such as scheme, hostname, path and query. A domain is only part of a typical website URL. Keep public URLs stable where practical and redirect deliberately when moving or renaming important pages.

91. Upgrade

A change to a newer version, larger allocation or different service. These are different kinds of upgrade with different risks. Identify compatibility, migration work and recovery options before proceeding; paying for more storage will not solve every processing or code problem.

92. VPS

A virtual private server provides an operating-system environment on shared underlying infrastructure with defined allocations. It offers more control than many shared-hosting plans. Management responsibilities, resource commitments and supported recovery tools depend on the actual product, not the abbreviation alone.

93. Version control

A method for recording changes to files and coordinating work over time. It supports review and rollback of tracked material. It does not automatically cover uploaded media, production databases or secrets, so it belongs alongside rather than in place of backup procedures.

94. Virtualisation

Creating virtual representations of computing resources, commonly virtual machines on physical hardware. It enables separation and flexible allocation. Performance and isolation depend on the technology and configuration; the existence of a virtual machine does not remove underlying physical dependencies.

95. Web server

The software answering web requests, and sometimes the machine running it. It may serve files directly or pass requests to an application. Diagnosing problems requires distinguishing the network, web server, application and database rather than calling every failure a hosting outage.

96. WordPress

An open-source content management system used for publishing and many other website functions through suitable extensions. Self-hosted WordPress requires compatible hosting and maintenance. Themes, plugins and custom code form part of the site’s operational responsibilities, not just its visual design.

97. Wildcard certificate

A certificate with wildcard hostname coverage, commonly matching one label such as *.example.com. That does not automatically cover the bare domain or arbitrarily deep names such as a.b.example.com. Check the actual names included instead of assuming one wildcard covers every possible hostname.

98. Zone file

A text representation of DNS records for a zone. DNS services may expose the same information through an interface or API. Exporting and reviewing the records before migration helps preserve website, email and verification services that share the domain.

99. Zero downtime

A goal or claim of no observed service interruption during a defined change or period. Achieving it requires appropriate architecture, deployment and measurement. It should not be presented as an inherent hosting feature that makes every possible failure impossible.

100. Zlib

A software library for data compression used by other applications and tools. Its presence is an implementation detail rather than a complete website-optimisation strategy. Keep software using it maintained and distinguish the library from a particular server’s HTTP compression settings.

Put the terminology to work

When a feature list uses one of these terms, turn it into a practical question. Ask what the backup covers, how a restore works, where the server actually runs or who applies patches. Record the answer against the application’s needs. Clear responsibilities and usable evidence matter more than the length of a technical feature list.

For a first website, identify the CMS, storage, email and maintenance you need, then compare a suitable hosting service. You do not need to administer every layer yourself, but you should know which organisation or person is responsible for it.

About the editorial team

Practical guidance from the telaHosting UK editorial desk, with a focus on clear explanations, useful next steps and traceable sources.

Our editorial approach

Your perspective

Join the discussion

Ask a question, share what worked, or add a useful detail.

Discussion

Comments

Loading discussion…

Join the conversation

Your email stays private. Contributions are reviewed before publication.

Keep it useful and respectful. Plain text only.0 / 2,000
Moderated discussion

Reply to the discussion

Write a reply

Your email stays private. Contributions are reviewed before publication.

Keep it useful and respectful. Plain text only.0 / 2,000
Moderated discussion

Keep exploring

More in this topic
Server hardware beside a globe showing the British Isles and Europe.Web hosting basics

UK or Overseas Hosting? Choose by Evidence, Cost and Support

A UK-facing website, a British supplier and a UK server are three different things. Compare performance, billing, support, security, search and payment arrangements, then use a verified entry-budget example to choose the hosting service that fits your business without relying on nationality or unsupported speed claims.

7 min readUK article
Processor and copper heatsinks with an amber signal on a circuit board.Web hosting basics

How Server Response Time Affects UK Website Visitors

Server response is part of the time a customer waits, but it is not the whole page experience. Explore ten practical effects, understand location and latency, and use a controlled investigation to identify whether the problem is the network, application, infrastructure or browser before recommending a fix.

6 min readUK article

Find a useful guide

Explore guides by topic, task or product.

Explore the journal