--- title: "Dedicated server checklist: what to ask before you sign | StreetHosting" description: "What a business should settle before buying a dedicated server: maintenance windows, reinstalls, IPMI, hardware replacement, backup, recovery and on-call." url: "https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business" type: "page" language: "en-US" --- Infrastructure · 11 min · Intermediate Published on Sep 15, 2026 · Updated on Sep 15, 2026 # What to settle before signing for a dedicated server Almost every dedicated server headache comes from something nobody asked before signing. This checklist gathers the questions a business needs to put to the provider, and what to document internally, so you do not find out the answer in the middle of an incident. By [Equipe StreetHosting](https://streethosting.com.br/en/autores#equipe-streethosting) · StreetHosting infrastructure and support team [Hardware and datacenter](https://streethosting.com.br/en/guides/topics/hardware) Summarize with: [](https://chat.openai.com/?q=Summarize%20the%20key%20points%20of%20this%20StreetHosting%20guide%3A%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business.%20Highlight%20the%20step-by-step%20instructions%2C%20the%20prerequisites%20and%20the%20most%20common%20mistakes. "ChatGPT") [](https://claude.ai/new?q=Summarize%20the%20key%20points%20of%20this%20StreetHosting%20guide%3A%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business.%20Highlight%20the%20step-by-step%20instructions%2C%20the%20prerequisites%20and%20the%20most%20common%20mistakes. "Claude") [](https://www.google.com/search?udm=50&aep=11&q=Summarize%20the%20key%20points%20of%20this%20StreetHosting%20guide%3A%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business.%20Highlight%20the%20step-by-step%20instructions%2C%20the%20prerequisites%20and%20the%20most%20common%20mistakes. "Google AI Mode") [](https://x.com/i/grok?text=Summarize%20the%20key%20points%20of%20this%20StreetHosting%20guide%3A%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business.%20Highlight%20the%20step-by-step%20instructions%2C%20the%20prerequisites%20and%20the%20most%20common%20mistakes. "Grok") [](https://www.perplexity.ai/search/new?q=Summarize%20the%20key%20points%20of%20this%20StreetHosting%20guide%3A%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business.%20Highlight%20the%20step-by-step%20instructions%2C%20the%20prerequisites%20and%20the%20most%20common%20mistakes. "Perplexity") Share: [](https://x.com/intent/tweet?text=Dedicated%20server%20checklist%3A%20what%20to%20ask%20before%20you%20sign&url=https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business "Share on X") [](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business "Share on Facebook") [](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business "Share on LinkedIn") [](https://wa.me/?text=Dedicated%20server%20checklist%3A%20what%20to%20ask%20before%20you%20sign%20https%3A%2F%2Fstreethosting.com.br%2Fen%2Fguides%2Finfrastructure%2Fdedicated-server-checklist-for-business "Share on WhatsApp") For agents: Copy as Markdown [.md](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business.md) In this guide 7 sections * [Why settle it before signing](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#por-que-alinhar-antes) * [Maintenance windows and reinstalls](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#manutencao-e-reinstalacao) * [Physical access, IPMI and console](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#acesso-fisico-e-ipmi) * [Hardware failure and replacement](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#hardware-e-substituicao) * [Whose backup, and the recovery plan](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#backup-e-recuperacao) * [On-call and escalation](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#plantao-e-escalonamento) * [What to document before signing](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#o-que-documentar) Quick answer Before **buying a dedicated server for your business**, settle seven points with the provider in writing: how the maintenance window works, who can request a reinstall, what access exists to the console and IPMI, what the procedure is when a component fails, who is responsible for backups, what the recovery plan looks like and who answers outside business hours. All of it should become a question in the pre-sales ticket, not an assumption. ## Why settle it before signing[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#por-que-alinhar-antes) Buying a dedicated server differs from buying a VPS in one fundamental way: there is a physical machine with your name on it. Disks age, power supplies fail, memory throws errors and the rack goes through maintenance. None of that is a sign of a bad provider. It is the reality of any hardware, and the difference between a calm operation and a panicked one lies in knowing in advance how each of those events is handled. The most common mistake in corporate purchasing is confusing the sales page with an operational agreement. The page describes the product. What your business needs is the answer to process questions, and those answers almost always exist, they just are not published. They show up when someone asks. The sections below are organized as questions to ask, not promises to wait for. Take this list to the pre-sales ticket, record what gets answered and keep the ticket number. That record is worth more on the day of an incident than any uptime figure advertised in bold. Ask these questions before choosing the plan, not after. Some answers change the configuration you should be buying, and switching machines after migrating costs far more than asking first. ## Maintenance windows and reinstalls[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#manutencao-e-reinstalacao) Scheduled maintenance on network, power and racks happens in every datacenter. What your business needs to know is how far in advance you are notified, through which channel and whether there is any way to negotiate the time when it collides with your business peak. An online store and a network of game servers peak at very different moments, and that information only reaches the provider if you tell them. The reinstall policy is the other side. Reinstalling the operating system wipes data, so the procedure needs a safeguard. Who at your company is authorized to request that operation. If someone from internal tech support opens the ticket, does the provider execute it directly or confirm with the account owner. When reinstalling is available as an automatic action in the control panel, anyone with panel access has the power to wipe everything, and that changes who should have a login. There is a third element in this conversation, which is firmware updates and infrastructure component updates. Security fixes for motherboard and disk controller firmware usually require a machine reboot. Ask how those updates are carried out, whether you are warned beforehand and whether you have any room to pick the moment. A company that runs month-end batch processing, for example, has two or three days a month when rebooting is especially bad, and the provider only knows that if you say so. * Ask how far in advance scheduled maintenance is announced and through which channel. * Tell the provider what your business's critical window is, with hours and days. * Confirm whether there is an emergency maintenance notice channel and who on your team receives it. * Define in writing who at the company can authorize a system reinstall. * Check whether the reinstall is self-service through the control panel and restrict who has that access. * Record whether the reinstall preserves any partition or wipes the whole disk. ## Physical access, IPMI and console[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#acesso-fisico-e-ipmi) At some point the server will not come up. A broken kernel after an update, a misconfigured firewall that locked out your own access, a disk that was not recognized at boot. In those moments SSH does not help, because it depends on a working system. What solves it is out-of-band access, which on most servers is IPMI or an equivalent remote console offered by the provider. Ask whether that access is handed directly to you or goes through a ticket. Both answers are legitimate and each has a practical consequence. Direct access means autonomy and recovery in minutes by your own team. Ticketed access means depending on support response time, which can be perfectly acceptable if that time is short and support works in the middle of the night. Physical access to the machine is a separate subject. A serious datacenter does not let just anyone into the room, and that is a quality, not a limitation. What matters is knowing how to request a physical intervention when one is needed, who performs it and how the request is authenticated. The guide on [what a data center tier means](https://streethosting.com.br/en/guides/infrastructure/data-center-tiers-explained) helps you understand what the infrastructure itself guarantees. Whatever the format, test that access in the first week, with the machine still empty and no rush. Finding out that the console credential never worked is far cheaper before production is sitting on top of it. Include that test in the server acceptance, along with the disk, memory and network checks, and record the result in the same document that holds the provider's other answers. * Confirm whether IPMI, remote KVM or an equivalent console is available. * Find out whether out-of-band access is direct to the customer or opened on request through a ticket. * Test that access in the first week, before you actually need it. * Ask how a physical intervention is requested and who can request it. * Document the authentication procedure for critical requests. ## Hardware failure and replacement[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#hardware-e-substituicao) This is the question almost nobody asks and everybody wishes they had. When a disk fails, what is the path. Does the provider detect it on its own through monitoring or wait for you to open a ticket. Does the swap happen with the machine running or does it require a shutdown. If it requires a shutdown, who picks the time. And most important: after the swap, who restores the data. It is worth asking the same about memory and the power supply. Each component has a different failure mode. Faulty memory can take the machine down intermittently for days before anyone understands the cause, which is why ECC memory shows up in configurations aimed at corporate environments, such as the XL Intel Mid Large plan with 384 GB of DDR4 ECC in the [dedicated servers](https://streethosting.com.br/en/dedicated) lineup. There is a second scenario, rarer and more serious, that deserves its own question: the case where the whole machine needs to be replaced. It can happen through multiple failures, end of life of the equipment or an infrastructure decision by the provider. In that scenario what matters is not the replacement lead time, it is how long your business takes to rebuild the environment from scratch on another machine from backup. If the internal answer is "we don't know", that is the biggest risk in the purchase, and it is not on the provider's side. | Event | Question to ask the provider | What your business needs to define | | ------------------------------- | -------------------------------------------------- | --------------------------------------------------------- | | Disk failure | How the failure is detected and how the swap works | Where the data will be restored from | | Memory failure | Is there a diagnosis before replacement | How much downtime is tolerable | | Power supply or network failure | Does redundancy cover this component | Who is notified and through which channel | | Machine will not boot | What out-of-band access exists | Who on the team performs the recovery | | Full machine replacement needed | How migration to another machine works | How long it takes to restore the environment from scratch | Do not accept or demand a magic replacement-time number without understanding the process behind it. A short lead time advertised without a clear procedure is worth less than a well-described procedure. Ask how it works and record the answer. ## Whose backup, and the recovery plan[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#backup-e-recuperacao) On an unmanaged dedicated server, the data is the customer's responsibility. That needs to be clear in writing before the first byte is copied to the machine. Many companies assume the provider keeps a copy by default and find out otherwise on the worst possible day. If there is a backup service you can contract, ask what it covers, how often it runs, where the copy is stored and how a restore is requested. Having a backup and having a recovery plan are different things. The backup is the file. The recovery plan is knowing, in order, what to do when the machine disappears: where the latest copy is, how long it takes to download, which environment it will be restored to, who performs it and how the end customer finds out. Without that written plan, the company improvises under pressure and the loss grows. Two numbers need to be decided internally and nobody can decide them for you. How long the operation can stay down and how much recent data can be lost. Those two values define the backup frequency and the kind of redundancy worth paying for. The [3-2-1 strategy](https://streethosting.com.br/en/guides/infrastructure/3-2-1-backup-strategy) is the starting point, and the [incident response plan](https://streethosting.com.br/en/guides/infrastructure/incident-response-plan-gaming-community) covers the communication side. * Confirm in writing who is responsible for backups under the plan you are buying. * Define where the off-server copy lives, away from the dedicated machine. * Decide how much downtime the operation can tolerate and record that number. * Decide how much recent data can be lost in an extreme case. * Test a full restore before considering the migration complete. * Repeat the restore test at a fixed interval, with a date on the calendar. ## On-call and escalation[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#plantao-e-escalonamento) A serious incident rarely happens at ten on a Tuesday morning. It happens in the early hours of a Saturday, during an event, with everyone asleep. The practical question is simple: who answers at that moment, on the provider's side and on your company's side. On the provider's side, confirm whether support runs 24 hours a day and in which language. Support in Portuguese matters more than it seems when the subject is technical and urgent, because describing a network symptom in another language, under pressure, adds noise and costs time. Also confirm which channels work in the middle of the night and which one is the official channel for incidents, because it is not always the same one used for sales. It is also worth understanding how the provider classifies a ticket. An information request and a total service outage cannot go into the same queue, and most providers have some form of prioritization. Ask how your team flags that something is a serious incident and avoid using that flag for things that are not, because its value depends on being taken seriously on both sides. On your company's side, define the escalation before you need it. Who gets called first, who steps in if that person does not answer within fifteen minutes, who has the authority to approve a decision that costs money at three in the morning, such as restoring yesterday's backup and losing the day's transactions. Without that definition, everyone waits for someone else to decide. * **Official incident channel:** which one it is, and whether it is monitored outside business hours. * **Support language:** technical, in Portuguese, in the middle of the night. * **First internal contact:** name, phone number and backup person. * **Decision authority:** who approves losing recent data to bring the service back faster. * **Communication with the end customer:** who writes it and where it gets published. A [status page](https://streethosting.com.br/en/guides/infrastructure/incident-status-page) handles this better than scattered messages. What to ask for in terms of support, in greater detail, is in [Support SLA: what to ask for](https://streethosting.com.br/en/guides/infrastructure/game-hosting-sla). ## What to document before signing[](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#o-que-documentar) Everything answered in the previous sections needs to become an internal document. Not a parallel contract, but a page your team can open at three in the morning and find the information without digging through old conversations. That document is what turns pre-sales answers into operational process. It is worth keeping that document off the dedicated server itself. It sounds obvious said like that, but it is common for recovery documentation to be hosted on exactly the machine you intend to recover. A team repository, a shared company space or even a printed copy kept by the person in charge will do. The criterion is simple: if the server vanished right now, this content stays accessible. * Pre-sales ticket number with the provider's answers recorded. * Machine identification, contracted plan and primary IP address. * Name of the account owner and of whoever can authorize destructive operations. * How to access the control panel and the remote console, and where the credentials are kept. * Where the backup is, how often it runs and when the last restore test was. * Internal escalation list with name, contact and backup person for each position. * Official channel for opening an incident with the provider outside business hours. * The business's critical window as communicated to the provider, with hours and days. * Date of the next review of this document. With that material ready, the purchase stops being a leap in the dark. You know what to expect, who to call and what to do when something goes off script. To size the machine itself and understand the price range of each configuration, see the guide on [dedicated server pricing and plans in Brazil](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-brazil-price), and for the architecture around it check the [StreetHosting infrastructure](https://streethosting.com.br/en/infraestructure) options. In this guide * [Why settle it before signing](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#por-que-alinhar-antes) * [Maintenance windows and reinstalls](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#manutencao-e-reinstalacao) * [Physical access, IPMI and console](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#acesso-fisico-e-ipmi) * [Hardware failure and replacement](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#hardware-e-substituicao) * [Whose backup, and the recovery plan](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#backup-e-recuperacao) * [On-call and escalation](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#plantao-e-escalonamento) * [What to document before signing](https://streethosting.com.br/en/guides/infrastructure/dedicated-server-checklist-for-business#o-que-documentar) ## Frequently asked questions What should I ask a provider before buying a dedicated server? Ask how the maintenance window works, who can request a reinstall, whether there is IPMI or remote console access, what the procedure is when a component fails, who is responsible for backups and how support works outside business hours. Ask for the answers in writing in the pre-sales ticket. Who is responsible for backups on a dedicated server? On an unmanaged dedicated server, backing up the data is the customer's responsibility unless a backup service is specifically contracted. Confirm this in writing before migrating any data and set up your off-server destination from day one. What is IPMI and why does it matter when signing up? IPMI is the out-of-band management interface on the server's motherboard. It lets you power the machine on and off, watch the boot screen and mount media even when the operating system is frozen. Without some equivalent access, any boot problem becomes a ticket that depends on the provider. Do I need an in-house team to run a dedicated server? You need someone responsible, even if it is an outside vendor. Bare metal requires an update routine, disk monitoring and a tested recovery plan. The provider takes care of the physical infrastructure and the network, not your application. How does a business evaluate support before signing? Open a pre-sales ticket with a real technical question outside business hours and measure how long the first useful reply takes. That tells you more about support than any sales page copy. Next step See dedicated servers Exclusive hardware in São Paulo with NVMe and Anti-DDoS. [See dedicated servers](https://streethosting.com.br/en/dedicated) [Explore the infrastructure Datacenter in São Paulo, gamer network and options for high-demand projects.](https://streethosting.com.br/en/infraestructure) ## Related guides [Infrastructure Beginner SLA and support in game hosting: what to demand Support that only opens a generic ticket at 9 a.m. while your season is on fire at 10 p.m. is not cheap: it has a hidden cost. This guide lists what to ask for before you pay for a year. 2 min Read guide](https://streethosting.com.br/en/guides/infrastructure/game-hosting-sla) [Infrastructure Beginner The 3-2-1 backup strategy for servers The 3-2-1 rule is one of the most reliable ways to reduce data loss after a technical failure or a human mistake. 2 min Read guide](https://streethosting.com.br/en/guides/infrastructure/3-2-1-backup-strategy) [Infrastructure Beginner What is uptime and what does a 99.9 percent SLA mean Every provider promises high availability, but few explain what that means in hours. Here is the math behind uptime and what a real SLA should guarantee. 3 min Read guide](https://streethosting.com.br/en/guides/infrastructure/what-is-uptime-and-sla) [← Back to the Guide Center](https://streethosting.com.br/en/guides)