Refurbished Server Offers - Configurations, Prices & Stock

Build a comparable shortlist of public refurbished server offers with the configuration actually listed: CPU count and total cores, installed memory, storage description, bays, rails, RAID, management and warranty. Keep missing components and out-of-stock offers visible instead of filling gaps with chassis maximums.

Start with a small run

Run with the prefilled input to observe up to five rows from the default public collection. No API key, login or proxy is required. Inspect the dataset plus SUMMARY and the readable RUN_REPORT before feeding observations into purchasing decisions.

Ready-configured server shortlist

Observe three listings with their actual installed configuration text.

{
  "collectionUrls": [
    "https://techmikeny.com/collections/servers"
  ],
  "productsPerSource": 3,
  "maxResults": 3
}

Known Dell certified-refurbished offer

Monitor one stable SKU, including its honest current stock status.

{
  "productUrls": [
    "https://techmikeny.com/products/dell-certified-refurbished-poweredge-r960-server-2-20ghz-80-core-64gb-600gb"
  ],
  "maxResults": 1
}

Dell R640 exact configuration

Read one NVMe server configuration without confusing supported maximums with installed RAM.

{
  "productUrls": [
    "https://techmikeny.com/products/dell-poweredge-r640-nvme-server-2x-gold-6154-3-00ghz-36-core-128gb-h730p-rails"
  ],
  "maxResults": 1
}

What you get

Field	Meaning
recordId / sku	Stable configured-offer identity from the public source.
manufacturer / serverModel / generation	Seller-specified make and model; null where absent.
cpuCount / totalCpuCores / cpuDescription	Installed CPU information; raw listing text is retained.
ramTotalGb / ramModuleCount / ramModuleCapacityGb	Installed memory from the listing, not the maximum RAM a chassis can support.
driveBayCount / installedStorageText	Bays and the actual drive/caddy description. A bay count does not mean drives are installed.
railsIncluded / raidController / managementText	Included components where the source explicitly says so; otherwise null.
condition / warrantyText	Literal seller terms, which must be checked before purchase.
configurationBasis	listed_configured_system, base_configuration or public_offer_configuration_unspecified. A starting/base price does not imply an upgraded system.
price / currency / available	Public asking price and source-reported stock; no inventory count or checkout guarantee.
url / sourceUrl / observedAt	Product link, exact observed endpoint and observation time.
comparisonStatus / changedFields / previous	Optional saved-observation comparison with the prior observed values.
priceChange / priceChangePercent	Same-currency difference from the prior saved observation; null without a comparable prior price.

Save a watchlist and compare later

Save your input as an Apify Task. Set comparePrevious: true and choose a snapshotName. The first complete run establishes a baseline and marks observations new; later runs mark them changed or unchanged. Set onlyChanges: true to omit unchanged rows. The same source URLs, selected window, availability filter and snapshot name share history. Different scopes stay separate.

Only complete, successfully saved observations advance history. Source failures, unreachable sources, output or time caps, spending limits and concurrent history conflicts do not overwrite the prior baseline. A saved row can be useful even when the overall run is partial; check SUMMARY. Missing items never prove removal or discontinuation. Concurrent tasks should use separate snapshot names.

Bounds and source behavior

The Actor reads Product JSON-LD already delivered in public TechMikeNY collection and product HTML. It does not use the custom server configurator, private APIs, customer accounts or checkout. When a collection contains more records than your selected window, later records are outside coverage. Unknown or aggregate-only offers are rejected rather than converted into an invented configured price.

• Select up to 10 exact product URLs and 3 collection URLs on the supported source host.
• Each collection observes a window of 1–20 products; maxResults saves 1–100 rows. Filtering applies after the window is selected.
• The default is five selected products and five saved rows, with a 90-second budget; the maximum budget is 120 seconds.
• Requests respect robots.txt, enforce body and time limits, reject redirects and stop on HTTP 401, 403 or 429. There is no bypass or repeated-until-success retry.
• This is a selected-window monitor, not a full catalog or market comparison. SUMMARY always distinguishes that scope from complete coverage of the selected input.

Configuration text can be incomplete or change. The Actor does not infer compatible CPU upgrades, maximum supported memory, installed-drive totals from bay counts, or CPU power from a power-supply wattage. Base/starting configurations remain labeled.

Pricing

The planned Actor price is $0.002 per saved result plus $0.00005 per Actor start. For example, five saved rows cost $0.01005 in Actor event charges, with any separate Apify platform usage charges shown by your plan. The live Store pricing is authoritative. Each result is charged once when saved; unchanged rows omitted with onlyChanges do not incur a result charge. A run can still incur the start event and platform usage when it saves no rows. Set a per-run maximum charge in Apify for a firm spending cap.

Export and use

Download CSV, Excel or JSON from the dataset, or use the Apify API and your saved Task in a procurement workflow. Always retain source URLs and observation times beside price comparisons. Confirm condition, warranty, compatibility, availability, tax and shipping with the seller before placing an order. Asking prices are not completed sales, fair value, future resale profit or guaranteed margins.

Support

Open an issue on this Actor with the run ID and a public source URL. Do not include credentials or private purchasing information. Explore Mayd It data tools and our quality notes.

Source consistency checks reject conflicting installed CPU/RAM values instead of choosing one. Explicit RAM totals written in TB use 1,024 GB per TB and must agree with the stated module count and module capacity when all are present. This RAM convention does not describe usable disk capacity.