Four-Hour Response Is a Weaker Promise Than Six-Hour Repair
HPE Tech Care Essential commits an engineer to attend your site within four hours. Tech Care Critical commits the hardware to be repaired within six. Read those as "4 hours" against "6 hours" and you will pick the weaker commitment believing it is the stronger one. The difference between attendance and repair is the difference between somebody arriving and your server running, and it is the single most misread line in HPE support quoting.
This page sets out what the base warranty already covers, what each Tech Care level actually commits to, where Complete Care is a different kind of product entirely, and the site-level question that decides whether any of it applies at your address.
What the warranty already gives you
Before adding anything, a new ProLiant is not unsupported. The DL380 Gen12 base warranty is 3-year parts, 3-year labour, 3-year onsite support with next business day response. Three things around that are worth knowing because they surprise people:
- Drives can carry a different term. HPE states plainly that hard drives have "either a one year or three-year warranty" and directs you to the specific drive QuickSpecs. A three-year server can contain one-year drives.
- Software support is 90 days, not three years. The base warranty covers hardware diagnosis and repair; software support and initial setup assistance run for 90 days from purchase.
- Customer Self Repair applies. Some parts are designated mandatory CSR — you are expected to fit them yourself, and if you decline and ask HPE to do it, travel and labour charges apply. Optional CSR parts can be replaced by HPE at no charge on request. For a lights-out site or a customer without hands on the ground, this matters more than the headline warranty length.
So the real question is never "warranty or no warranty". It is whether next-business-day onsite is fast enough for what the machine does, and whether you have someone at the site who can fit a part.
The three Tech Care levels
| Basic | Essential | Critical | |
|---|---|---|---|
| Coverage window | 9 hours per day, local HPE standard business hours, excluding weekends and HPE holidays | 24×7 including HPE holidays | 24×7 including HPE holidays |
| Phone response | 2-hour callback from a product specialist | 15 minutes for severity 1; 1 hour for severity 2–3 | 15 minutes for severity 1; 1 hour for severity 2–3 |
| Hardware commitment | Next business day onsite response | 4-hour onsite attendance | 6-hour hardware call-to-repair (severity 1–2) |
| Outage management | — | — | Yes, for severity 1 business-impacting situations |
All three levels include phone and expert chat access, general technical guidance, help with HPE proprietary service tools, replacement parts and materials, firmware updates for selected products, the service portal, and remote problem diagnosis. The differences are the clock and the nature of the promise.
Response is not repair
This deserves its own section because it decides the money.
- Onsite response or attendance means an engineer arrives within the window. Nothing is committed about what happens next. If the part is wrong, or the fault needs a second visit, or the depot has to ship overnight, the clock has already been satisfied.
- Call-to-repair means the covered hardware is returned to operating condition within the window. Getting the engineer there is HPE's logistics problem, not the boundary of the commitment.
Which means Critical's six hours is a materially stronger guarantee than Essential's four, despite the larger number. If your own SLA to a customer or to the business is written in terms of restored service, only call-to-repair maps onto it. Buying four-hour attendance and reporting it upward as a four-hour recovery guarantee is a commitment you cannot keep.
The corollary is also worth stating plainly: Critical is not needed everywhere. For a node in a cluster designed to lose a node, next-business-day is entirely rational — the workload has already failed over and the repair is housekeeping. Buying Critical across a fleet where the architecture provides the resilience is paying twice for the same availability. The tiering should follow your failure domains, not your org chart, and that argument is easiest to make when node count and failure domain were designed together — see HCI node sizing.
Complete Care is a different kind of product
Tech Care is per-device operational support. HPE Complete Care is an account-level engagement, described by HPE as its "most comprehensive support solution" for customers wanting "an advanced, tailored service experience". What you are buying is people and process rather than a response clock:
- An assigned service delivery team and a service experience manager as a single point of contact
- A technical advisor who is a deep product specialist, with advisory sessions and collaborative incident support
- Proactive elements — security assessment and insights, tailored alerts, automatic case creation, firmware analysis
- Technical lifecycle planning and platform profiling
It is modular and scoped to an estate rather than attached to a part number, so it is not directly comparable with a Tech Care line on a quote. The honest test is whether you would otherwise employ someone to do that coordination. If your team already runs firmware baselines, tracks lifecycle and manages incidents competently, Complete Care duplicates them. If you are running a multi-country estate with a small team, it can be cheaper than the headcount — but price it against headcount, not against Tech Care.
Where Foundation Care still appears
HPE moved its support portfolio to Tech Care through 2021 — ProLiant DL servers from April, some storage from May, and full enablement across products by July. The old levels map across directly:
| Previous offering | Tech Care level |
|---|---|
| Foundation Care NBD · Proactive Care NBD | Basic |
| Foundation Care 24×7 · Proactive Care 24×7 | Essential |
| Foundation Care CTR · Proactive Care CTR | Critical |
Foundation Care has not vanished. HPE's own store still carries Foundation Care SKUs, but look at what they are attached to: older generations such as DL360 Gen10, legacy platforms such as BladeSystem c-Class, parts of the networking portfolio, and post-warranty coverage. Current ProLiant generations attach Tech Care.
The practical consequence is on quotes for mixed estates. A quote covering Gen10 machines alongside Gen12 may legitimately carry both Foundation Care and Tech Care lines. That is not an error, but you should be able to say which line covers which machines, and confirm the two levels are actually equivalent rather than assuming the names line up.
The question that decides whether any of it applies
Response-time commitments are not universal. They depend on site eligibility, which is generally tied to how far the installation address sits from an HPE support location and parts depot. A four-hour or six-hour commitment that is available in one city may simply not be offered at a site a few hundred kilometres away, and the answer differs country by country.
For an estate spread across Hong Kong, the Gulf, Africa or the CIS, this is the first thing to settle, not the last. Before a support level goes into a contract or an internal SLA:
- Get the committed level confirmed per installation address, in writing — not per country, and not per order.
- Ask where the parts depot serving that address is, and what is stocked there.
- Ask what the commitment degrades to if the site is outside the eligible radius.
- Check whether customs clearance sits inside or outside the repair clock. On cross-border sites this can dominate everything else — the same issue covered in export and customs handling.
Where a site cannot get the level you need, the answer is usually architectural rather than contractual: hold spares locally, or design the cluster to survive the outage. That is a decision to make at purchase, while the spares can still be added to the same order.
What to specify
- State the outcome you need — "service restored within X hours" or "next business day is acceptable" — rather than a service level name. This alone prevents the response-versus-repair confusion.
- Say which machines are which. Tiering per role is normal and cheaper than one level across the fleet.
- Give every installation address, so eligibility is checked before the contract, not after an incident.
- Say whether anyone is on site who can fit a Customer Self Repair part.
- Specify the term and whether it should match the intended hardware life. A three-year support term on a machine you plan to run six years leaves a gap you will pay more to fill later — the post-warranty renewal is priced quite differently.
Support should be on the quote from the start for the same reason management licensing should be: it is a per-device line that multiplies across a fleet, and it is negotiable during a bid in a way it is not afterwards. The sequence is in buying 20–60 servers as one project, and the parallel licensing line is in iLO and Compute Ops Management licensing.
Get support quoted per machine role, not one level across the fleet
Send this and we price it — no questions back:
- Server count, and which machines are redundant cluster nodes versus single points of failure
- The outcome you need — "service restored within X hours" or "next business day is acceptable" — rather than a service level name
- Every installation address, so we check eligibility before it goes in a contract
- Whether anyone is on site who can fit a Customer Self Repair part
- How many years the hardware is expected to run, so the support term does not end first
You get firm pricing, availability and delivered lead time within one business day.
No specification yet? Send quantity, the workload it runs (or committed vCPU and RAM), storage capacity, destination country and target date — we come back with a specified configuration and a price, not a list of questions.
Get support levels priced per site Prefer email? sales@haink.org
We confirm the committed level per installation address before it goes on the quote, and we will tell you where a site cannot get the level you asked for rather than quoting it and discovering it during an incident.
Frequently asked questions
What does a new HPE server come with by default?
On the DL380 Gen12, three-year parts, three-year labour and three-year onsite support with next business day response. Note two limits: drives may carry a one-year or three-year warranty depending on the specific drive, and software support with initial setup assistance runs for 90 days rather than three years.
What is the difference between Tech Care Essential and Critical?
Essential commits to four-hour onsite attendance — an engineer arrives. Critical commits to six-hour hardware call-to-repair for severity 1 and 2 — the hardware is returned to operating condition — and adds outage management for business-impacting severity 1 incidents. Critical is the stronger commitment despite the larger number of hours, because it guarantees an outcome rather than an arrival.
Is Foundation Care discontinued?
Not discontinued, but it is no longer what current ProLiant generations attach. HPE moved the portfolio to Tech Care through 2021, and Foundation Care SKUs now appear mainly against older generations, legacy platforms, parts of networking, and post-warranty coverage. Foundation Care NBD, 24×7 and CTR map to Tech Care Basic, Essential and Critical respectively.
Do I need Complete Care?
Only if you want the assigned team, the service experience manager, lifecycle planning and proactive analysis — it is an account engagement, not a faster response clock. Price it against the cost of the headcount that would otherwise do that coordination. If your team already runs firmware baselines and incident management well, it duplicates them.
Does a four-hour response guarantee my server is fixed in four hours?
No. Four-hour onsite attendance guarantees an engineer arrives within four hours. Nothing after arrival is committed. If your internal or customer-facing SLA is written in terms of restored service, the only level that maps onto it is call-to-repair.
Is the response time the same at every site?
No. Commitments depend on site eligibility, which is generally tied to distance from an HPE support location and parts depot, and availability differs by country. Confirm the committed level per installation address in writing before it goes into a contract. Where a site cannot get the level you need, local spares or a more resilient cluster design is the practical answer.
Should every server have the same support level?
Usually not. Match the level to the failure domain: a node in a cluster designed to lose a node rarely justifies call-to-repair, while a single non-redundant machine running something important justifies it easily. Uniform high-tier support across a resilient architecture means paying twice for availability you already engineered.
Related
- How to choose HPE ProLiant servers and storage — platform selection and host sizing
- Aruba vs Juniper under HPE — neither line is being killed, so which to standardise on
- HPE four-socket scale-up — DL580 Gen12 against two DL380, and why 4U buys only memory
- iLO and Compute Ops Management licensing — the other per-device line that multiplies across a fleet
- Sourcing HPE · part numbers decoded — verifying warranty entitlement before payment
- Lenovo warranty vs Premier Support — the same response-versus-repair distinction on another badge
- HCI node sizing — why the support tier should follow the failure domain
- Alletra explained · GreenLake vs owning — where support is bundled rather than quoted separately
- Buying 20–60 servers as one project · export and customs · secondary market
- HPE stock, lead times and pricing — current availability and firm quotes
Sources
- HPE — Tech Care Service, operational support service data sheet (Basic: "9 hours per day during local HPE standard business hours", "2-hour callback from product specialist", "Next business day on-site response for covered hardware". Essential: "24 hours per day, 7 days per week including HPE holidays", 15-minute severity 1 callback, "4-hour on-site attendance for covered hardware". Critical: "6-hour hardware call-to-repair" for severity 1–2 plus outage management)
- HPE — ProLiant Compute DL380 Gen12 QuickSpecs ("3-year parts, 3-year labor, 3-year onsite support with next business day response"; "Hard drives have either a one year or three-year warranty"; 90 days software support and initial setup; mandatory and optional Customer Self Repair parts)
- HPE — Complete Care Service (assigned service delivery team; "your service experience manager that acts as your single point of contact"; technical advisor; security assessment and insights; auto case creation; firmware analysis; technical lifecycle planning and platform profiling)
- HPE — Tech Care Service (current service portfolio and level definitions)