80+ Akamai Interview Questions and Answers (2027)
90 Akamai interview questions with answers: Property Manager, caching, Fast Purge, EdgeWorkers, Terraform, App & API Protector, Bot Manager, Prolexic.
Akamai interview questions test how well you can deliver and protect web traffic on Akamai's edge: configure properties, rules and caching in Property Manager, purge content, test on the staging network, automate with the Akamai CLI and Terraform provider, write EdgeWorkers, and secure sites with App & API Protector, Bot Manager and Prolexic. This page has 90 questions with direct answers, grouped by experience level.
Product names and behaviour were checked against Akamai TechDocs and Akamai product pages in November 2026. Some security product docs on techdocs.akamai.com sit behind a customer login, so those answers stay at the level Akamai publishes openly.
Key takeaways
- Freshers are asked how a request reaches an Akamai edge server: CNAME to an edge hostname, property rules, caching, CP codes and the staging and production networks.
- Candidates with 2 to 5 years of experience should read a rule tree, use Pragma debug headers, purge by URL, CP code or cache tag, and manage properties with the Terraform provider or Property Manager API.
- Security roles focus on App & API Protector (the product that succeeded Kona Site Defender), WAF tuning, rate controls, Bot Manager, API security and Prolexic DDoS protection.
- Edge compute questions cover EdgeWorkers event handlers (
onClientRequest,onOriginRequest,onOriginResponse,onClientResponse,responseProvider) and EdgeKV. - Linode is now Akamai Cloud. Know its core services (Linodes, LKE, NodeBalancers, Object Storage) in case the role spans CDN and cloud compute.
- Scenario questions on stale content, a falling cache hit ratio, WAF false positives and live DDoS attacks carry the most weight for experienced roles.
How to use this guide
Each answer opens with a short reply you could give in an interview, then adds the detail an interviewer may probe next. For general DevOps rounds around the CDN topics, see our 200+ DevOps interview questions guide.
| Your experience | Read first | Then practise |
|---|---|---|
| Fresher or career switcher | Questions 1 to 28 | Trace a request with dig and curl to see the CNAME chain and cache headers |
| 2 to 5 years | Questions 29 to 55 | Write a property in Terraform, activate it on staging and test with Pragma headers |
| Security engineer | Questions 56 to 72 | Explain how you would tune a WAF policy from alert to deny |
| Senior, lead or architect | Questions 73 to 80 | Sketch delivery and security for a global site, including origin protection |
| Everyone with work experience | Questions 81 to 90 (scenarios) | Explain each scenario aloud: symptom, check, fix, prevention |
Akamai interview questions for freshers
1. What does Akamai do?
Akamai runs a large distributed network of edge servers that deliver and protect web content, video, downloads and APIs close to users. Its portfolio covers content delivery (Ion, Adaptive Media Delivery, Download Delivery), security (App & API Protector, Bot Manager, Prolexic, API Security, Guardicore Segmentation) and cloud computing through Akamai Cloud, the platform formerly known as Linode.
2. What is a CDN, and why use one?
A content delivery network caches and serves content from servers near the user instead of from one origin. Users get lower latency, the origin handles far fewer requests, and the CDN absorbs traffic spikes. Akamai also uses its edge to inspect traffic, so the same layer that caches content can block attacks before they reach your servers. Our post on CDN integrations in cloud-native deployments covers where a CDN fits in a delivery pipeline.
3. How does a user's request reach an Akamai edge server?
- The user looks up
www.example.com. - Your DNS has a CNAME from
www.example.comto an Akamai edge hostname such aswww.example.com.edgekey.net. - Akamai's mapping system answers with the IP of an edge server suited to that user.
- The edge server finds the property for that hostname, applies its rules, and serves from cache or goes forward to the origin.
4. What is an edge hostname?
An edge hostname is the Akamai-generated CNAME target that puts your hostname on the Akamai network. Hostnames under edgekey.net are used for Enhanced TLS, and edgesuite.net hostnames are used for Standard TLS or non-secure delivery. You create the edge hostname, then point your own hostname at it in DNS.
5. What is an origin?
The origin is the server where your content lives, for example a load balancer, web server, cloud storage bucket or Akamai NetStorage. Edge servers fetch from it when content is not in cache or has expired. The origin hostname is a DNS name for that server, often something like origin-www.example.com, which should not be the same as the public hostname.
6. What is a property in Property Manager?
A property is the container for your delivery configuration for a site, app, download area or stream. It applies a set of rules to a set of hostnames. Only one property can apply to a given hostname at a time.
7. What is a property version?
A property version is a snapshot of the configuration. You edit a new version while another version stays active, and once you activate a version it is locked. To change it again you create a new version. This gives you a history and a quick way back to a known good state.
8. What are rules, criteria and behaviors?
| Term | Meaning | Example |
|---|---|---|
| Rule | A block of criteria and behaviors; rules can nest up to five levels under the default rule | "Static content" |
| Criteria (match) | The IF part: when the rule applies | File extension is css, js or png |
| Behavior | The THEN part: what the edge does | Cache for 7 days |
9. What is the default rule?
The default rule is the top-level rule that applies to every request for the property. It usually holds the origin behavior, the CP code and baseline settings. More specific rules sit underneath it as children and override or add to those settings for matching requests.
10. What is a CP code?
A content provider (CP) code is an identifier Akamai uses to track and report traffic. Every property needs at least one CP code behavior in its default rule. Teams create extra CP codes to split reporting and billing, for example one for images and one for API traffic, and they can purge by CP code.
11. What are the staging and production networks?
They are the two Akamai networks you can activate a property version on. Staging is a smaller set of edge servers for testing; production serves real users. You activate on staging first, test by pointing your machine at the staging edge, then activate the same version on production.
12. How do you test a property on staging?
Resolve the staging form of your edge hostname and send your test traffic to that IP while keeping the real Host header and SNI.
dig +short www.example.com.edgekey-staging.net
curl -sI https://www.example.com/ --resolve www.example.com:443:<staging-ip>
For browser testing, add the staging IP for your hostname to your local hosts file, then remove it when you finish.
13. How do you roll back a bad activation?
Activate the previous known good version again. Akamai also supports fast fallback, which lets you roll a production activation back to the previous version quickly within a short window after activation. You can also cancel an activation shortly after requesting it.
14. Which Akamai delivery products should you know?
| Product | Typical use |
|---|---|
| Ion | Websites and web apps, with performance optimisations |
| Dynamic Site Accelerator | Dynamic, mostly uncacheable sites that need faster origin routes |
| Adaptive Media Delivery | Video streaming (HLS, DASH) |
| Download Delivery | Large file downloads such as installers and game patches |
| Image & Video Manager | Automatic image and video optimisation per device |
15. What are a cache hit and a cache miss?
A hit means the edge served the object from its cache. A miss means it had to fetch it from a parent tier or the origin. The time to live (TTL) decides how long an object stays fresh. The offload or hit ratio, the share of requests served without going to origin, is one of the main numbers Akamai teams watch.
16. Which options does the caching behavior offer?
The main choices are to cache with a fixed max-age, honor the origin's Cache-Control and/or Expires headers, not store the object (no-store), or bypass the cache for that request. Fixed TTLs suit static assets; honoring origin headers suits teams that control caching in the application.
17. What is downstream cacheability?
It controls the caching headers Akamai sends to browsers and other downstream caches, separately from how long Akamai itself caches the object. You might cache an HTML page for 5 minutes at the edge but tell browsers not to cache it, so a purge at Akamai takes effect for users immediately.
18. What is a cache key, and how can you change it?
The cache key is what the edge uses to decide whether two requests are for the same object. By default it includes the hostname, path and query string. The Cache ID Modification behavior lets you include or exclude query parameters, cookies, headers or variables, for example ignoring tracking parameters like utm_source so they do not split the cache.
19. What is Fast Purge, and how do invalidate and delete differ?
Fast Purge removes or refreshes content across the edge within seconds, through Control Center, the Fast Purge API or the purge CLI. Invalidate marks objects stale, so the next request revalidates with the origin; delete removes them, so the next request fetches the full object. Akamai recommends invalidate, because it is gentler on the origin.
20. What can you purge by?
By URL, by CP code, or by cache tag. Cache tags come from an Edge-Cache-Tag response header your origin sends, so you can tag every page that shows product 42 and purge them all with one call when the price changes.
akamai install purge
akamai purge invalidate https://www.example.com/css/site.css
akamai purge invalidate --cpcode 12345
akamai purge invalidate --tag product-42
Akamai's guidance is that purging should not replace sensible TTLs.
21. What is Tiered Distribution?
Tiered Distribution adds a parent tier between edge servers and your origin. On a miss, an edge server asks a parent first, and only the parent goes to origin. This raises offload and protects the origin, which matters most for content with a long tail of rarely requested objects.
22. What is SureRoute?
SureRoute finds faster or more reliable paths across the internet between edge servers and your origin by racing alternate routes with a test object. It helps dynamic, uncacheable requests and gives a failover path when the direct route degrades.
23. What is NetStorage?
NetStorage is Akamai's cloud storage built for use as an origin. Teams use it for large files, video segments and static sites, uploading through supported protocols or the API, and then deliver it through a property.
24. What is Edge DNS?
Edge DNS is Akamai's authoritative DNS service. It answers DNS queries for your zones from Akamai's network, which keeps DNS available under heavy query loads and DDoS attacks. It also supports an AKAMAICDN record type for pointing a zone apex at an edge hostname, where a normal CNAME is not allowed.
25. What is Global Traffic Management (GTM)?
GTM is DNS-based load balancing. It checks the health of your data centres with liveness tests and answers DNS queries with the best available data centre, based on rules such as failover, weighted distribution or performance. Use it for active-active or active-passive setups across regions or clouds.
26. How does Akamai handle TLS certificates?
For Enhanced TLS you get certificates through the Certificate Provisioning System (CPS) or use Default DV certificates managed by Akamai in Property Manager. Standard TLS uses Akamai's shared certificate model. Enhanced TLS is the usual choice for production sites with their own domain, and for compliance needs such as PCI.
27. What is Akamai Control Center?
Control Center is Akamai's web portal. You use it to manage properties, certificates, security configurations, DNS zones, reports, users and API clients. Most teams use it alongside APIs, the CLI and Terraform rather than only clicking through it.
28. What is Akamai Cloud?
Akamai Cloud is Akamai's cloud computing platform, formerly sold under the Linode brand after Akamai bought Linode in 2022. Its services include Linodes (virtual machines, including GPU types), Linode Kubernetes Engine (LKE), NodeBalancers, Object Storage, Block Storage, Managed Databases, Cloud Firewall and VPC. The Linode name still appears in product and tool names such as the Linode CLI.
Akamai interview questions for 2 to 5 years of experience
29. What does a property rule tree look like in JSON?
The Property Manager API returns rules as a JSON tree. This trimmed example sets an origin and CP code in the default rule and caches static files for seven days:
{
"rules": {
"name": "default",
"behaviors": [
{ "name": "origin", "options": { "originType": "CUSTOMER", "hostname": "origin-www.example.com", "forwardHostHeader": "REQUEST_HOST_HEADER" } },
{ "name": "cpCode", "options": { "value": { "id": 12345 } } }
],
"children": [
{
"name": "Static content",
"criteria": [
{ "name": "fileExtension", "options": { "matchOperator": "IS_ONE_OF", "values": ["css", "js", "png", "jpg", "svg"], "matchCaseSensitive": false } }
],
"behaviors": [
{ "name": "caching", "options": { "behavior": "MAX_AGE", "mustRevalidate": false, "ttl": "7d" } }
]
}
]
}
}
30. What is a rule format?
A rule format is a dated, versioned set of behaviors and criteria that a rule tree uses. Pinning a rule format means features do not change under you; setting latest picks up new behavior versions automatically. Production teams usually pin and upgrade on purpose.
31. What is the Property Manager API (PAPI)?
PAPI is the REST API behind Property Manager. With it you can create properties, read and update rule trees, create versions, manage hostnames and activate on staging or production. The Terraform provider and CLI packages call these APIs for you.
32. How do you authenticate to Akamai APIs?
With EdgeGrid, which signs each request with HMAC-SHA-256 credentials from an API client you create in Identity and Access Management. The credentials go in an .edgerc file in your home directory:
[default]
client_secret = REPLACE_ME
host = akab-xxxxxxxx.luna.akamaiapis.net
access_token = akab-xxxxxxxx
client_token = akab-xxxxxxxx
Give each person and each pipeline its own API client with only the permissions it needs, and keep the file out of Git.
33. What is the Akamai CLI?
The Akamai CLI is a command-line tool you extend with packages, such as purge, property-manager, edgeworkers and appsec. You install it from GitHub releases or with brew install akamai, and packages with akamai install <package>. Packages read credentials from .edgerc, and the --edgerc and --section flags choose a different file or section.
34. How do you manage a property with the Akamai Terraform provider?
Use the akamai/akamai provider with EdgeGrid credentials, then define the CP code, edge hostname, property and activation as resources.
terraform {
required_providers {
akamai = {
source = "akamai/akamai"
}
}
}
provider "akamai" {
edgerc = "~/.edgerc"
config_section = "default"
}
resource "akamai_edge_hostname" "web" {
contract_id = var.contract_id
group_id = var.group_id
product_id = var.product_id
edge_hostname = "www.example.com.edgekey.net"
ip_behavior = "IPV6_COMPLIANCE"
certificate = var.cps_enrollment_id
}
resource "akamai_property" "web" {
name = "www.example.com"
contract_id = var.contract_id
group_id = var.group_id
product_id = var.product_id
rule_format = var.rule_format
rules = data.akamai_property_rules_template.web.json
hostnames {
cname_from = "www.example.com"
cname_to = akamai_edge_hostname.web.edge_hostname
cert_provisioning_type = "CPS_MANAGED"
}
}
resource "akamai_property_activation" "staging" {
property_id = akamai_property.web.id
version = akamai_property.web.latest_version
network = "STAGING"
contact = ["[email protected]"]
}
Pin the provider version in real projects. The same Terraform habits apply as on any cloud; see our Terraform interview questions.
35. How would you promote Akamai changes through a pipeline?
Keep rules as JSON snippets or a rules template in Git. On a pull request, run terraform plan and lint the JSON. After merge, apply and activate on staging, run automated checks against staging (status codes, cache headers, redirects), then activate the same version on production with a manual approval. Never edit production by hand in Control Center, or Terraform state and reality will drift.
36. What are Pragma debug headers?
They are request headers that ask the edge to return diagnostic headers. They show whether the object was cached, the cache key and whether it is cacheable.
curl -sI https://www.example.com/css/site.css \
-H "Pragma: akamai-x-cache-on, akamai-x-cache-remote-on, akamai-x-check-cacheable, akamai-x-get-cache-key, akamai-x-get-true-cache-key"
The response then includes headers such as X-Cache, X-Cache-Remote, X-Check-Cacheable, X-Cache-Key and X-True-Cache-Key.
37. How do you read the X-Cache header?
It shows the cache result at the edge server. TCP_HIT and TCP_MEM_HIT mean served from cache; TCP_MISS means fetched from a parent or origin; TCP_REFRESH_HIT means the object was stale and was revalidated with the origin. X-Cache-Remote shows the same for the parent tier, which tells you whether the parent or the origin answered.
38. A user sends you an error page with "Reference #18.xxxx". What do you do?
That reference string identifies the edge server and the request. Paste it into the error translation tool in Control Center (Edge Diagnostics) to see the reason, for example an origin timeout, a connection refused by origin, a WAF deny or a property rule that returned 403. This is usually the fastest first step in any Akamai incident.
39. Why might a cache hit ratio be low?
- The origin sends
Cache-Control: no-storeorprivateand the property honors origin headers. - Query strings or tracking parameters split the cache key.
- Cookies or
Varyheaders make responses uncacheable. - TTLs are very short for content that rarely changes.
- Frequent broad purges by CP code empty the cache.
- Long-tail content without Tiered Distribution.
40. How should query strings be handled in the cache key?
Include only the parameters that change the response. Exclude marketing and tracking parameters, or whitelist the few meaningful ones, using Cache ID Modification. Mistakes go both ways: excluding a real parameter such as ?page=2 serves the wrong content, while including every parameter lowers the hit ratio.
41. How do cookies affect caching?
A response that sets cookies or depends on them may be personalised, so caching it can leak one user's content to another. Cache only responses that are the same for everyone, bypass cache for logged-in paths, and if you must cache with cookies present, make sure the cache key includes what makes the response differ.
42. What is Site Failover?
Site Failover is a behavior that tells the edge what to do when the origin fails or times out: serve an alternate origin, redirect to a backup hostname, or serve a static fallback page from NetStorage. Combined with GTM, it keeps a site partly available during origin outages.
43. How do you protect the origin so only Akamai can reach it?
Restrict origin firewalls to Akamai's published origin-facing IP ranges, for example with Site Shield or the Origin IP Access Control List behavior, and add a shared secret header or mutual TLS between edge and origin so a request with a spoofed source cannot get through. Without this, attackers can bypass the CDN and its WAF by hitting the origin directly.
44. What is mPulse?
mPulse is Akamai's real user monitoring (RUM) product. A script in the page measures what real users experience, such as page load time, Core Web Vitals and errors, broken down by page group, browser, country and network. It answers "is the site fast for users", which server-side metrics cannot.
45. What is DataStream 2?
DataStream 2 streams edge request logs in near real time to destinations such as Amazon S3, Azure Storage, Google Cloud Storage, Splunk, Datadog or an HTTPS endpoint. Teams use it to analyse cache status, errors, latency and suspicious clients in their own observability stack.
46. What is Image & Video Manager?
It optimises images and short videos at the edge. Based on the requesting device and browser, it resizes, compresses and converts images to efficient formats, so one master image in the origin produces many derivative versions without manual work.
47. What are Cloudlets?
Cloudlets are small, policy-driven edge applications managed separately from the property rules, so non-CDN teams can change them quickly. Examples include Edge Redirector (bulk redirects), Phased Release (send a percentage of users to a new origin), Application Load Balancer, Request Control (allow or deny by request attributes), Visitor Prioritization (waiting room) and Forward Rewrite.
48. What are EdgeWorkers?
EdgeWorkers lets you run your own JavaScript on Akamai edge servers. You can change requests and responses, make routing decisions, personalise content or build responses at the edge. You deploy code through the API, CLI or Control Center, and attach it to traffic with the EdgeWorkers behavior in a property.
49. Which EdgeWorkers event handlers exist?
| Event handler | When it runs |
|---|---|
onClientRequest | For every request, before the cache lookup |
onBotSegmentAvailable | After onClientRequest, only with Bot Manager Premier, to read the bot score |
onOriginRequest | Just before the request goes to origin (not on cache hits) |
onOriginResponse | When the origin response arrives (not on cache hits) |
onClientResponse | Before the response goes to the client |
responseProvider | When the response is not in cache; acts as a surrogate origin that builds the response |
50. What goes into an EdgeWorkers code bundle?
A gzip-compressed tar file containing main.js, which exports the event handler functions, and bundle.json, which holds the version and description, plus any modules you import.
{
"edgeworker-version": "1.0.3",
"description": "Add security headers and geo redirect"
}
51. How do you deploy an EdgeWorker from the command line?
akamai install edgeworkers
tar -czvf ew-bundle.tgz main.js bundle.json
akamai edgeworkers upload --bundle ew-bundle.tgz 12345
akamai edgeworkers activate 12345 STAGING 1.0.3
Test on staging, then activate the same version on production. Each EdgeWorker ID is tied to a resource tier that sets limits on CPU time, memory and sub-requests.
52. What is EdgeKV?
EdgeKV is Akamai's distributed key-value store for EdgeWorkers. It suits data that is read often and changed rarely, such as feature flags, redirect maps or per-country settings. Writes take time to spread across the network, so do not use it like a transactional database.
53. Why would an EdgeWorker fail or be skipped?
Common causes are exceeding the resource tier's CPU or memory limits, an uncaught exception, a sub-request that times out, the EdgeWorkers behavior not matching the request in the property, or a version that was never activated on that network. Check the EdgeWorkers debug headers on staging and the error reports, and keep handlers small and fast.
54. How does Akamai compare with CloudFront and Cloudflare?
| Point | Akamai | Amazon CloudFront | Cloudflare |
|---|---|---|---|
| Configuration | Properties with rule trees, PAPI, Terraform | Distributions and behaviors in AWS | Zones and rules in a dashboard or API |
| Edge code | EdgeWorkers (JavaScript) | CloudFront Functions, Lambda@Edge | Workers |
| Security | App & API Protector, Bot Manager, Prolexic | AWS WAF, Shield | WAF, bot management, DDoS protection |
| Typical buyers | Large enterprises, media, banks | Teams already on AWS | Wide range, from small sites to enterprises |
For the other two in detail, see our Cloudflare interview questions and S3 with CloudFront posts.
55. How would you cache API responses on Akamai?
Cache only GET responses that are the same for all callers, such as product catalogues or public reference data, with short TTLs and cache tags for targeted purges. Bypass cache for authenticated, user-specific or write requests. Make sure the cache key includes the parameters and headers that change the response, and keep downstream caching off if clients must see purges immediately.
Akamai security interview questions
56. Which Akamai security products should you know?
| Product | Protects against |
|---|---|
| App & API Protector | Web and API attacks (WAF), layer 7 DDoS, with rate controls and bot controls |
| Bot Manager | Automated traffic such as scrapers, credential stuffing and inventory hoarding |
| Account Protector | Account takeover and fraudulent logins |
| Content Protector | Content scraping |
| API Security | API discovery, posture and abuse detection |
| Prolexic | Network-layer DDoS against data centres and IP ranges |
| Edge DNS | DNS outages and DNS-based DDoS |
| Guardicore Segmentation | Lateral movement inside data centres and clouds |
| Enterprise Application Access | Zero trust access to internal apps without a VPN |
57. What is App & API Protector?
App & API Protector is Akamai's web application and API protection product, running on the edge. Akamai lists capabilities including a WAF with OWASP Top 10 coverage, the Adaptive Security Engine, a Behavioral DDoS Engine for layer 7 attacks, API discovery, bot controls and a malware protection module. App & API Protector Hybrid extends WAF protection to on-premises, hybrid cloud and multi-CDN environments.
58. What happened to Kona Site Defender?
Kona Site Defender was Akamai's earlier WAF product. Akamai now sells App & API Protector as its web application and API protection product, and its Kona Site Defender product page redirects there. Older contracts and documentation may still use the Kona name, so recognise it, but describe current setups in App & API Protector terms.
59. How are App & API Protector settings organised?
A security configuration holds one or more security policies. Match targets decide which hostnames and paths each policy protects. Inside a policy you set WAF protections, rate controls, IP and geo firewall rules, custom rules and other protections. Like properties, security configurations are versioned and activated on staging and then production.
60. What is the Adaptive Security Engine, and what are alert and deny?
The Adaptive Security Engine is Akamai's self-tuning WAF rule engine; Akamai describes it as learning attack patterns and pushing updated protections automatically. Rules and attack groups (such as SQL injection, cross-site scripting and command injection) can be set to alert, which only logs, or deny, which blocks. New policies usually start in alert, are tuned with exceptions, then move to deny.
61. What are rate controls?
Rate controls (rate policies) count requests from a client to matching paths over a time window and take an action, alert or deny, when a threshold is crossed. Typical uses are login endpoints, search and checkout APIs. Set thresholds from real traffic data, and exclude known partners or monitoring IPs.
62. What are IP/Geo firewall rules and client lists?
IP/Geo firewall rules allow or block traffic by IP, CIDR range, country or ASN. Client lists are reusable, centrally managed lists of IPs, countries or other identifiers that several policies can reference, so updating one list updates every policy that uses it.
63. When would you write a custom rule?
When the built-in protections do not cover an application-specific risk, for example blocking requests to an old admin path, requiring a header on an internal API, or blocking a pattern seen in an attack that the managed rules miss. Keep custom rules few and documented, because each one is something your team now maintains.
64. What is the penalty box?
The penalty box applies an action to a client for a period of time after it triggers a deny. Instead of checking each new request from an attacker in full, the edge applies the penalty action to that client directly, which reduces load during attacks.
65. How do you handle a WAF false positive?
Find the request in security event logs or with the reference ID, identify the rule or attack group that fired and the part of the request that matched (often a parameter or cookie value). Add a narrow exception for that rule on that path or parameter, test on staging, then activate. Avoid switching a whole attack group to alert to fix one form field.
66. What does Bot Manager do?
Bot Manager identifies automated traffic and lets you choose how to respond. It keeps a directory of known bots (such as search engines and monitoring services), lets you define your own bot categories, and scores unknown clients from 0 (human) to 100 (bot) using behaviour analysis and browser fingerprinting. Responses range from monitoring through challenges to mitigation such as denying, slowing or serving alternate content.
67. How do Account Protector and Content Protector differ from Bot Manager?
Bot Manager asks "is this a bot, and what should we do about it". Account Protector focuses on login and account flows, scoring whether the person using an account is its owner to stop account takeover. Content Protector targets scrapers that copy content and pricing. They are separate products that work alongside each other.
68. What is Akamai API Security?
It discovers APIs from traffic (including ones teams forgot to document), checks their security posture, and detects abuse and attacks on API logic, such as data scraping through valid calls. Akamai built this product line after acquiring Noname Security in 2024. It complements App & API Protector, which blocks attacks inline.
69. What is Prolexic?
Prolexic is Akamai's DDoS protection for whole networks and data centres, not just websites behind the CDN. Traffic for your IP ranges is routed through Akamai scrubbing centres, which drop attack traffic and forward clean traffic to you. It protects any internet-facing service, including non-HTTP ones.
70. What Prolexic deployment options exist?
| Option | How it works | Fits |
|---|---|---|
| Prolexic Routed (GRE) | You advertise your prefix through Akamai with BGP; clean traffic returns over GRE tunnels | Owners of at least an IPv4 /24 or IPv6 /48 |
| Prolexic Connect | Add-on with dedicated links between Akamai scrubbing centres and your data centre | High bandwidth, predictable latency |
| Prolexic IP Protect | Anycast protection for smaller or fragmented address space | Less than a /24, or cloud-hosted IPs |
Each can run always-on (all traffic through Prolexic all the time, the most common choice) or on-demand (routed only during an attack).
71. What is the SOCC?
The Security Operations Command Center is Akamai's team of security analysts. Depending on your service level, they monitor attacks, help mitigate DDoS events, and tune protections with you during incidents. Know how to open an incident with them before you need to.
72. How does Akamai support zero trust?
Enterprise Application Access gives users access to specific internal apps based on identity and device posture, without putting them on the network through a VPN. Guardicore Segmentation controls traffic between workloads inside data centres and clouds, so an attacker who gets in cannot move freely. Together they cover user-to-app and workload-to-workload access.
Senior Akamai interview questions
73. How would you design delivery and security for a global e-commerce site on Akamai?
- Ion property with Enhanced TLS, Tiered Distribution and Image & Video Manager for product images.
- Long TTLs plus cache tags for catalogue pages, bypass for cart, checkout and account paths.
- App & API Protector in deny mode with rate controls on login, search and checkout; Bot Manager against scalpers and scrapers.
- Origin locked to Akamai with Site Shield or IP ACLs, plus GTM across two regions and Site Failover.
- Everything in Terraform, activated on staging first, with DataStream 2 and mPulse for visibility.
74. What do you consider for a multi-CDN setup?
Traffic steering (DNS-based, often with GTM or a third-party steering service), consistent cache keys and TTLs across CDNs, purging every CDN on deploy, the same WAF posture everywhere (App & API Protector Hybrid targets this), certificate management on each CDN, and origin access lists that include every provider. Compare with the edge compute model in our Fastly interview questions.
75. How would you govern Akamai configuration changes in a large company?
Treat CDN and security configuration as code: Terraform or PAPI-driven pipelines, peer review on every change, staging activation with automated tests, a named approver for production, and least-privilege API clients per pipeline. Restrict Control Center write access so emergency changes are rare and later copied back into code.
76. How do you measure whether an Akamai setup is working?
Track offload (edge hit ratio and origin requests), origin error rates and response times, edge error rates by status code, real user metrics from mPulse, and security metrics such as denied attacks and false positive reports. Set targets for each and review them after every major change.
77. How do you deliver video at scale on Akamai?
Use Adaptive Media Delivery with HLS or DASH, short segments cached at the edge, manifests with shorter TTLs than segments, and an origin designed for segment fetches (often NetStorage or a cloud origin behind a parent tier). Monitor start-up time and rebuffering, and use token authentication for paid content.
78. How would you migrate a site from another CDN to Akamai?
Inventory the current rules (redirects, caching, headers, WAF rules), map each to Akamai behaviors and protections, and build the property in Terraform. Test on staging with real traffic patterns, run App & API Protector in alert mode first, lower DNS TTLs before the cutover, switch hostnames gradually, and keep the old CDN ready to take traffic back.
79. How do you keep EdgeWorkers code safe to deploy?
Store it in Git with unit tests, bundle it in CI, activate on staging and run tests against the staging edge, then promote the same version to production. Keep handlers short to stay within resource limits, log and fail open where the business allows it, and roll back by activating the previous version.
80. When would you choose Akamai Cloud compute alongside the CDN?
When workloads benefit from running close to Akamai's network and you want simple, predictable cloud building blocks: origins for media, API backends, Kubernetes on LKE, or GPU instances for inference. Large enterprise workloads with deep managed-service needs may still fit better on a hyperscaler, with Akamai in front for delivery and security.
Scenario-based Akamai interview questions
81. Users still see the old version of a page after a deploy. What do you check?
Check the response headers with Pragma debug headers to see whether the edge or the browser is serving the old copy. If the edge is, purge the affected URLs or cache tags with invalidate. If browsers are, look at downstream cacheability and use versioned file names for CSS and JS. Long term, add a purge step to the deploy pipeline.
82. The cache hit ratio dropped from 90% to 60% after a release. How do you investigate?
Compare cache keys and headers for the same URL before and after. Usual suspects: the app started sending Cache-Control: private or new cookies, a new query parameter or tracking tag splits the cache key, or a deploy job now purges a whole CP code. DataStream 2 logs grouped by cache status and path show where misses rose.
83. The WAF blocks some legitimate checkout requests. What do you do?
Get reference IDs or timestamps from affected users, find the events in security logs, and identify the rule and the matched field. Add a narrow exception for that rule on the checkout path and field, test on staging, and activate. Tell the business what changed, and watch for attacks targeting that exception.
84. The origin is overloaded during a big sale even though Akamai is in front. Why?
Requests are reaching the origin: dynamic pages with no caching, cache-busting query strings, bots hammering search or inventory APIs, or attackers bypassing Akamai and hitting the origin IP. Raise offload with micro-caching where safe, turn on Tiered Distribution, apply rate controls and Bot Manager, lock the origin to Akamai, and consider Visitor Prioritization as a waiting room.
85. Users report errors and send you a reference number. What is your process?
Translate the reference in Edge Diagnostics to get the cause. Origin timeouts or connection errors point at the origin or network; a 403 with a security reason points at the WAF or a firewall rule; a rule-generated error points at property logic. Then check whether it affects one region, one path or everything, and act accordingly.
86. The site shows certificate errors in browsers. What could cause it?
An expired or not-yet-renewed certificate, a hostname missing from the certificate's SAN list, a new hostname added to the property before its certificate was deployed, or DNS pointing to the wrong edge hostname type. Check the certificate served by the edge, then the CPS enrollment or Default DV status. Prevent it with monitoring on certificate expiry dates.
87. An activation succeeded, but the change is not visible. Why?
Possible reasons: you activated on staging, not production; the content is still cached under the old TTL; the change sits in a rule whose criteria do not match the request; another property handles that hostname; or your test machine still points to staging through the hosts file. Check with debug headers and confirm which property version served the request.
88. An EdgeWorker starts failing in production after a traffic increase. How do you handle it?
Check EdgeWorkers error reports for resource limit errors. Heavier traffic exposes slow sub-requests or large payload processing. Short term, roll back or make the code fail open; then optimise (cache sub-request results, cut work per request, move data to EdgeKV) or move to a higher resource tier if the work is justified.
89. A volumetric DDoS attack hits your data centre IP range directly. What happens with Prolexic?
With always-on Prolexic, traffic already flows through scrubbing centres, which drop the attack and forward clean traffic. With on-demand, your network team (or Akamai, if you bought a facilitated service) routes your prefixes to Prolexic through BGP when the attack starts. Open an incident with the SOCC, confirm mitigation, and review afterwards whether always-on would be safer.
90. The business wants example.com (the zone apex) on Akamai, but DNS will not allow a CNAME there. What are the options?
If the zone is on Akamai Edge DNS, use the AKAMAICDN record type to point the apex at the edge hostname. Otherwise use your DNS provider's apex alias feature if it supports CNAME-like records, or redirect example.com to www.example.com and put only www on Akamai.
How to prepare for an Akamai interview
- Learn the request path cold: DNS CNAME, edge hostname, property, rules, cache, origin.
- Practise reading cache headers with Pragma debug headers on a site you can test.
- Write a small property and activation in Terraform, even if you cannot apply it, and explain each resource.
- For security roles, explain how you would move a policy from alert to deny without blocking customers.
- Practise the scenario questions aloud using symptom, check, fix and prevention.
For more practice, see our scenario-based Akamai interview questions. DevOps Training Institute runs classroom batches in Kharadi, Pune and live online classes across India. The DevSecOps course covers the WAF, DDoS and pipeline security topics behind the security questions, and the Terraform course covers the infrastructure-as-code skills used with the Akamai provider.
FAQs
Is Kona Site Defender still asked about in Akamai interviews?
Sometimes, because older environments and documents use the name. Akamai now sells App & API Protector as its WAF and API protection product, and the Kona Site Defender page redirects to it. Answer in App & API Protector terms and mention the older name if the interviewer uses it.
What is the difference between Akamai staging and production networks?
Both run your property configuration, but staging is a smaller set of edge servers used for testing. You activate a version on staging, point your test machine at it, and then activate the same version on production for real users.
What is the difference between invalidate and delete in Fast Purge?
Invalidate marks content stale, so the next request revalidates with the origin. Delete removes it, so the next request fetches the full object from origin. Akamai recommends invalidate in most cases because it is lighter on the origin.
Do I need to know JavaScript for Akamai roles?
For EdgeWorkers work, yes. For CDN configuration and security roles, JSON rule trees, HTTP, DNS, TLS and Terraform matter more. Basic JavaScript reading skills still help when you review edge code.
Is Linode the same as Akamai Cloud?
Yes. Akamai bought Linode in 2022 and now calls the platform Akamai Cloud. Product names such as Linodes, Linode Kubernetes Engine and the Linode CLI keep the old brand.
How do Prolexic and App & API Protector differ?
App & API Protector protects websites and APIs delivered through Akamai at the application layer, including layer 7 DDoS. Prolexic protects whole IP ranges and data centres from network-layer DDoS by routing their traffic through Akamai scrubbing centres.
Which Akamai automation tools should a DevOps engineer learn?
The Akamai Terraform provider, the Akamai CLI with its purge, property-manager and edgeworkers packages, and the Property Manager and Fast Purge APIs with EdgeGrid authentication. Together they let you manage delivery and security as code.
How many Akamai interview questions should I practise?
Be able to answer questions 1 to 28 without notes, read a rule tree and the Terraform example, and talk through at least six scenario questions in detail. That covers most CDN and DevOps interview rounds.
What's Your Reaction?
Like
0
Dislike
0
Love
0
Funny
0
Angry
0
Sad
0
Wow
0