What a news site asks of hosting that a brochure site does not
News website hosting in India needs three things: a server in or near India, a cache you can purge the moment a story publishes, and enough CPU for a spike when a link goes viral. A shared plan carries a small desk; a VPS is the safer base once traffic is daily.
A brochure site changes twice a year. A news site changes twenty times a day, in bursts, usually between 9am and 11pm when readers are awake and reporters are publishing. Hosting has to survive that rhythm rather than a monthly average.
- Bursty writes: ten stories in an hour, then nothing for six.
- Spiky reads: a story picked up by a large account, then a peak ten times normal.
- Heavy pages: photos, embeds, ad tags and comment widgets on every article.
- Fresh URLs: category and tag archives that must update without a full re-crawl.
Five checks before you choose a plan
Judge a plan on these five, in this order. The first two decide whether the site feels fast; the last three decide whether it survives an election night.
| Check | What to look for | Why it matters for news |
|---|---|---|
| Server location | A data centre in India, or a CDN with Indian edge nodes | Time to first byte for a reader in Delhi or Patna |
| Cache purge | Manual purge from the panel, or an API the CMS can call | A corrected headline must go live in seconds, not in an hour |
| PHP workers | Concurrent processes, not just "unlimited" disk | Twenty simultaneous article views can exhaust a two-worker plan |
| Backups | Daily, off-server, restorable by you | A bad plugin deploy at 11pm should not cost the archive |
| Upgrade path | A VPS or cloud step with the same panel | Traffic grows in steps, and a migration mid-spike is expensive |
Storage is not on the list. A text-and-photo news site uses far less disk than a shop; it is CPU and concurrent connections that run out first.
Server location and Indian readers
Distance sets latency. A server in Mumbai or Delhi answers an Indian reader in tens of milliseconds; the same page served from Frankfurt adds a round trip that no plugin can remove.
Test it rather than assume it. Run a page speed test against the live article from an Indian location, read the server response time, and compare it with the time to first byte for a cached page. If the cached figure is under 200 ms and the uncached figure is over a second, the bottleneck is PHP and the database, not the network.
For a site with regional editions, a CDN in front of an Indian origin is the usual shape: the origin writes, the CDN serves, and a purge call clears the affected URLs within seconds of publishing.
A cache that clears when you publish
Page caching is what makes a news site affordable. It is also what publishes yesterday's headline. Both halves have to be configured.
- Turn on page caching in the host's stack or in a caching plugin, and set a long TTL for article pages, which rarely change after publication.
- Exclude the homepage, the category archives and any live blog from long TTLs, or purge them on publish.
- Connect the CMS to a purge hook: most caching plugins call the host's purge API on save, and most CDNs accept a URL purge by API.
- Publish a test story, then load the homepage in a private window. You know it worked when the new headline appears on the first load, without a hard refresh.
The failure to watch for is the opposite of a stale page: a cache that purges on every comment or every admin-ajax call, which throws away the benefit. Purge on publish and on edit, not on every request.
Surviving the spike
Spikes arrive without warning, usually as a link from a large account or a WhatsApp forward. What fails first is rarely bandwidth.
- PHP worker limit: requests queue, then time out, and the browser reports a 508 or a blank page.
- Database connections: un-cached archive pages open a new connection each, and MySQL refuses the surplus.
- Memory: a page builder plus three ad tags can exceed a 256 MB PHP limit on a popular article.
- Disk I/O: a backup that starts mid-spike competes with the database for the same disk.
The defence is cached HTML served before PHP runs, images on a CDN, and a backup window at 3am rather than 9pm. If the site is already cached at the edge, a spike is mostly a bandwidth event, and bandwidth is the cheapest resource on any plan.
Google News, sitemaps and what hosting actually affects
Hosting does not buy inclusion in Google News; Publisher Center decides that on the strength of the publication, not the server. What hosting affects is whether the crawler can read the site quickly and reliably.
- HTTPS on every URL, with no mixed-content warnings on article pages.
- A news sitemap that lists the last two days of stories, submitted and kept current.
- Stable article URLs; changing the permalink structure after launch resets every indexed link.
- Server response under a second on an uncached article, so a crawl of 200 new URLs finishes in minutes.
Register the publication in Publisher Center, keep the sitemap valid, and treat hosting as the floor rather than the strategy. A fast, stable server helps a crawl; it cannot make a thin site eligible.
Shared hosting or a VPS for a news desk
The line is publishing volume, not ambition. A desk publishing five stories a day with modest traffic fits a shared plan. A portal publishing fifty, with video and a live blog, does not.
| Shape of the site | What fits | Price (₹/mo) |
|---|---|---|
| One desk, a few stories a day, cached pages | Shared hosting, L-Basic | 300 |
| Several sections, steady daily traffic, email on the domain | Shared hosting, L-Unlimited | 400 |
| Heavy plugins, many concurrent editors | Shared hosting, L-Developer | 700 |
| Live blogs, video, ad stack, predictable spikes | VPS or cloud server | 5,900 for the entry cloud configuration |
Shared prices are per month before 18% GST. The cloud entry configuration is 2 vCPU with 8 GB RAM and 100 GB storage, also before GST, and it is the point where you stop sharing CPU with strangers.
What WebJi® provides a publisher
WebJi® has hosted Indian websites since 2005, from a single office in Jaipur, with about 99% of clients served online. Plans bill in INR with a GST invoice, and the entry cloud server is listed on the cloud hosting page.
Free migration is available subject to terms and approval, so a move off a slow host can be scheduled rather than rushed. Support is by email at support@webji.services, 24/7 in Hindi and English, with an average first reply under 60 minutes. The office phone is a business-hours line, not a support queue.
Frequently Asked Questions
Which hosting is best for a news website in India?
A plan with an Indian data centre, a cache you can purge on publish, and enough PHP workers for your busiest hour. Start on shared hosting if you publish a handful of stories a day, and move to a VPS once you run live blogs, video or a heavy ad stack.
How much traffic can shared hosting handle for a news site?
With page caching on, a shared plan serves thousands of article views a day without strain, because cached HTML never reaches PHP. The limit appears on uncached requests: search, admin screens, comment posting and the homepage during a purge. Those are what exhaust workers first.
How do news sites handle traffic spikes?
Serve cached pages before PHP runs, put images on a CDN, and keep database-heavy pages such as archives out of the hot path. Schedule backups away from peak hours. A spike then becomes a bandwidth event rather than a CPU queue, and bandwidth is the cheapest resource on any plan.
Do news websites need a VPS?
Not at the start. A VPS earns its cost when you publish continuously, run a live blog, host video, or carry an ad stack that needs memory. The signal is a pattern of 508 or 503 errors during your busiest hour even with caching on.
Does hosting affect inclusion in Google News?
Only indirectly. Publisher Center decides eligibility on the publication itself; hosting decides whether Googlebot can crawl 200 new URLs quickly. HTTPS, a current news sitemap and stable URLs matter more than the server brand, though slow responses do waste crawl budget.
What should I do the day before a big traffic event?
Warm the cache on the pages you expect to trend, move the backup window, confirm the purge hook works, and check that images are served from the CDN. Then publish a test story and load the homepage in a private window to prove the cache clears.