tech

2023-02-01 · open the original ↗

• C# developers no longer have to switch to Python for machine learning. Korzynski says that around 2020 he had to choose between C# and machine learning and went to Python, but now does both: ML.NET has matured since 2017, and C# wrappers or ports exist for TensorFlow, Keras, NumPy, Pandas and PyTorch. NumPy.NET wraps the real Python library with all its features, while NumSharp is a native port with about 98% of them.
• ONNX lets data scientists and application developers each use their own tools. He asks his modelers for an ONNX file, built in whatever framework they prefer, and runs it in C# without Python dependencies, in ML.NET, or hosted in Azure Machine Learning or AWS SageMaker; he describes it as a common format for a neural network's weighted graph, created in about 2017 by Microsoft and Facebook. His reason is that Python's ecosystem suits modeling but its tooling for dependency injection, project structure and UIs does not suit enterprise applications.
• GPU inference comes from adding a single NuGet package. Some of his document extraction models are large enough to need a GPU even at prediction time, as can vision models like YOLO; he says the ONNX runtime's GPU package brings CUDA and optimizers such as TensorRT without setting up CUDA himself.
• Building the model is the easy part; data and integration are hard. Korzynski says getting well-labeled data and fitting a model into a C# codebase are the difficult stages, and a cloud-hosted model called over REST simplifies integration; latency matters less than expected because most workloads are not real-time, as with a 5,000-page document where OCR alone takes at least about a second per page.
• Deploying many models multiplies infrastructure beyond an ordinary app. On premises, he says, an application with 50 models could need 50 Docker containers, several instances of each under Kubernetes, and custom routing between them; his company digitizes microfilm and microfiche, and a project he just finished processed about 2 billion documents, requiring models run across Hadoop. He says some companies chase ML as a buzzword when they do not need it.

2023-01-26 · open the original ↗

• Capture should be frictionless, but it is not hoarding. Miller presents Tiago Forte's CODE method (Capture, Organize, Distill, Express) from the book Building a Second Brain, which he taught as a Codemash workshop; at capture the point is to get an idea from email, articles or podcasts into a notes app before it evaporates, though he now asks whether he will really benefit from it, having hoarded in the past.
• Organize notes for actionability, not by subject. Forte's PARA sorts everything into Projects (a goal and an end date), Areas of responsibility (an ongoing quality standard such as health or finances), Resources (reference and hobby material) and Archives (kept but out of the way); Miller says projects usually stem from areas, the same item can fall in different buckets depending on your role, and he has applied the scheme beyond his notes app to OneDrive, Google Drive and Outlook.
• Distill in layers, each time you return to a note. Under Forte's "progressive summarization", the first read bolds whatever stands out, the next pass reads only the bold and highlights within it, and a later pass adds your own summary at the top; Miller says only the material that keeps energizing you gets the later passes, and each one saves your future self time.
• Build output from your own stored pieces, choosing them before ordering them. Miller cites Forte's aim of starting no project he is not already 80% done with from prior "intermediate packets", and his advice not to go to Google when it is time to write; Forte's "archipelago of ideas" splits the work into selecting the pieces and then sequencing them in a separate pass, which Miller says he finds hard to do at once.
• Software can resurface what you read and capture podcast moments. Miller says Readwise, a paid service, pulls in Kindle highlights and shows five a day by default, and he was on a roughly 300-day review streak at recording; Snipd, a free podcast app, clips and transcribes the moment around a triple-tap on his earbuds and sends it to Readwise and Notion (he uses Notion's free tier, with Evernote kept as an archive).

2023-01-25 · open the original ↗

• A team should agree one word per meaning and record it. Ritchie, a software architect, says using update, change and modify interchangeably creates ambiguity, so a team might fix "change" to mean replacing an entity and "modify" to mean altering its attributes; he wants decisions kept somewhere such as SharePoint or GitHub, in the style of architectural decision records, with a domain vocabulary that defines each term, and for teams of teams a weekly or bi-weekly meeting to air them.
• An "and" in a method name signals it does too much. Ritchie says a long name is fine if it is the correct term in context, but a name that needs "and" usually means the method should be split, which then gives shorter names.
• Start with placeholder names and rename as understanding grows. He warns against analysis paralysis over perfect names: "test method one" is fine to begin with, so long as names are refactored, or things split into several, as the team's understanding of the problem improves.
• An interface name should describe a behavior, usually as an adjective. Rather than prefixing a class name with "I", Ritchie suggests turning the verb into an adjective, his example being "transmit" into "transmissive", because the interface should say what something does, not just provide a point for injection.
• Code should use the business's own language, not a technical one. Ritchie draws on domain-driven design's "ubiquitous language": one vocabulary shared with the business avoids mapping between technical and business terms, which he says introduces ambiguity; understanding the technical vision narrows the words needed, and since customers often cannot say what they want, names will change as feedback comes in. He collects his material in a GitHub repo, which he says holds far more than his hour-long talk.

2023-01-25 · open the original ↗

• Bulk operations replace the one-at-a-time user form at scale. Under Azure Active Directory's Users blade, Bulk operations offers bulk create, bulk invite and bulk delete; Giard says the web form is fine for a few users but not practical for hundreds or thousands. All three work the same three steps: download a template, edit it (and save it, which the page does not say), then upload it.
• The CSV template has a required version line and marked required columns. The create template ("user create template.csv") opens with "version:v1.0" on the first line, which must be present; the second row holds the column headers, and columns marked "required" (name, username, initial password, block sign-in yes/no) must be filled or the row fails. Optional fields such as first and last name, job title, department and location can be added later, though Giard notes they matter if you base conditional access on those properties.
• Upload results are reported per row, with a failure reason. After submitting, a status link shows success and failure counts and, in each row's details, a failure reason beyond the unhelpful "bad request". His demo hit three errors: a username whose domain was not one of the tenant's verified domains (contoso.com and microsoft.com were rejected; only the tenant's initial domain was valid), a leading space before an email address, and a blank line at the end of the file. One row can succeed while another fails, so he removed the user already created before re-uploading.
• Bulk invite and bulk delete need far fewer columns. The invite template uses the same version line and header format, with only the email address and redirection URL required. The delete template needs only the username, which Giard suggests for removing many accounts at once, such as after divesting a company or dropping a contractor firm.
• Giard expects the CSV to come from an automated process. He suggests a nightly batch that exports from another system and transforms the data into the template, then checking the status report for rows that failed. He advises keeping the file as CSV rather than converting it to Excel, and renaming and archiving each file as a history of what was attempted.

2023-01-24 · open the original ↗

• Pi-hole is the easiest project and the one he uses most. The host says it runs in under 30 minutes as a network-wide ad blocker and DNS server; his dashboard shows it blocking over 30% of outbound queries, and he values it most on his IoT network against devices such as a Roku phoning home. He suggests pairing it with a filtering upstream DNS such as Cisco's FamilyShield for malware and adult content.
• PiVPN makes a home VPN server simple to stand up. It builds an OpenVPN or WireGuard server for remote access to home files and servers or to use the home connection while travelling, and generates QR codes to set up clients. The host runs it on his own home network and credits its maintainers with keeping it current and easy.
• OctoPrint is a must-have for anyone with a 3D printer. At its simplest it monitors prints and sends jobs over the network instead of by SD card; plugins add temperature monitors, touchscreen control and Octolapse time-lapses. The host has no tutorial of his own but says the OctoPi image flashes straight from the official Raspberry Pi Imager.
• Home Assistant is the deepest project and can automate almost anything networked. The open-source server auto-detects smart switches, bulbs, thermostats, vacuums and cameras; the host runs his office lighting from one Stream Deck button, his fish-tank lights and air pump, and his 3D printer's smart power strip on it. He has made no videos on it and points to dedicated channels.
• His newest build is a scheduled internet speed-test appliance graphed in Grafana. Built to test T-Mobile Home Internet, it layers the Ookla Speedtest CLI on Raspberry Pi OS, writes timestamped results to InfluxDB and graphs them in Grafana, and can be dropped into any network. He says the Grafana install can monitor other data too, and plans a video on it.

2023-01-24 · open the original ↗

• Security groups carry permissions; Microsoft 365 groups carry messages. Created from Azure Active Directory > Groups > New group, a security group lets permissions assigned to the group flow automatically to its users or devices. Giard describes a Microsoft 365 group as what used to be a distribution list, so one message reaches everyone in it.
• A group can optionally be made assignable to Azure AD roles. Answering yes reveals a roles section for picking which built-in roles the group gets; Giard leaves it at no for the demo.
• Membership is either assigned explicitly or computed by a rule. Assigned, the default, lets you pick users and even other groups as members, plus an owner (Giard says only one owner can be added). Dynamic user and dynamic device membership instead use expressions joined with "and" (meet all criteria) or "or" (meet either), such as city equals Chicago and a given department, or device OS type equals iOS.
• Members can be added after the group exists. Giard creates a security group with three assigned members, finds it by search in the groups list, and adds more people through its Members page.

2023-01-23 · open the original ↗

• Inviting brings an outside account in, unlike creating an internal user. Building on creating a user in episode 136, Giard uses Azure Active Directory

  Users > New user > "Invite external user", filling in a name, email address and first and last name. An optional personal message goes into the invitation email, and the user can be pre-placed in groups or a role other than the default user role.

• Block sign-in holds an invited user until they should start. Giard's example is a contractor who does not begin until the first of next month; he sets it to no so access starts immediately.
• Optional profile fields pay off in conditional access policies. Location, job title, department, company name and manager are not required, but Giard says they let policies filter on things like users in the United States.
• Accepting the emailed invitation grants sign-in, not permissions. The email carries an "Accept invitation" link and a consent page; afterwards he signs in at portal.azure.com with his own address but can see almost nothing until an administrator grants access to applications.

2023-01-22 · open the original ↗

• Security defaults are a fast baseline, not complete security. Giard calls them "not perfect" but a good way to get started in a hurry. They are switched on under Azure Active Directory > Properties > "Manage security defaults" at the bottom of the blade, by choosing yes and saving.
• Security defaults cannot coexist with conditional access policies. Any existing conditional access policies must be disabled first, and the security defaults dialog warns you if any remain.
• The defaults force MFA registration and block legacy protocols. Every user must register for multi-factor authentication before using applications; older protocols such as SMTP and POP3 and certain proprietary ones are refused; and MFA is required for high-security areas, his examples being the Azure portal and the Azure CLI.
• Turning them off asks only for a reason. Setting it to no prompts for why, then save. Giard expects people to do this when they want the extra flexibility of conditional access policies.

2023-01-21 · open the original ↗

• Conditional access policies cannot run alongside security defaults. Giard says security defaults must be deactivated before conditional access policies can be used. Policies are created in Azure Active Directory under the Security tab, Conditional access blade, by creating a new policy and giving it a name.
• Assignments say which users and applications a policy covers. Users can be all users, specific users or groups, guests and external users, or directory roles; applications can be all cloud apps or selected ones, with a filter builder that requires custom attributes to be configured. Nearly every choice has include and exclude, so a policy can apply to a selection or to everyone except it, and Giard warns against a policy that locks yourself out.
• Conditions narrow when the policy fires, by risk, device and location. User risk flags a user on a leaked-credential list or behaving strangely; sign-in risk flags unusual activity such as signing in from two distant locations at once, each set to high, medium or low. Other conditions are device platform (e.g. apply to Linux, or to everything except Windows), locations (trusted, selected, or a list of forbidden countries or hostile IP ranges), client app (browser versus mobile app) and a device filter built from expressions on attributes such as device ID or manufacturer.
• Grant either blocks access or adds a required step first. Giard calls the required steps "speed bumps": multi-factor authentication, a password change, or accepting terms of service before access is allowed. The Session settings set how often the step repeats, for example a sign-in frequency of 4 hours, or not again within the same browser session, since he says repeating MFA on every page request would be an inconvenience.
• A policy can be created on, off, or report-only. Off lets an administrator configure a policy ahead of time, such as for the start of a quarter, and switch it on later. Report-only, which Giard finds useful, records who would match the policy without enforcing it, to learn whether it is needed at all.

2023-01-20 · open the original ↗

• The scam email genuinely comes from PayPal, so authentication checks pass. The host's two emails came from service@paypal.com with SPF, DKIM and DMARC all passing, because the scammer sends a real invoice through PayPal's invoicing system; the "View and pay invoice" link and every other link point to real PayPal pages. He guesses the scammer knew from some past leak that he was a Coinbase customer.
• Logging into PayPal directly is the surest way to expose it. The email claimed a $479 USD payment to Coinbase and warned that cancellation after 48 hours would not be refunded; logging into the account without clicking any link shows no such transaction. The host reads the deadline as pressure to act.
• The scam pays out through a careless payer or a phone call. The host suspects it hopes someone, such as an accounting department paying invoices without scrutiny, simply pays it, or calls the "help desk" number, where scammers try to take over the caller's computer. He urges sharing the warning with anyone who pays PayPal invoices.
• The phone numbers were the weakest tell, and both were already dead. One email gave an 888 toll-free number and the other a 213 Los Angeles number, odd for a Silicon Valley company but easy to miss. Calling both from a fresh VoIP number a day after receiving them, the host was hung up on immediately; he infers scammers rotate numbers once one is flagged.

2023-01-10 · open the original ↗

• The Orange Pi 5 is a powerful Raspberry Pi alternative you can actually buy. It has a Rockchip 8-core 64-bit CPU, 4, 8, 16 or 32 GB of RAM, a Mali-G610 GPU rated for 8K at 60 fps, gigabit Ethernet, full-size HDMI and an M.2 slot. The host doubts most buyers need 32 GB on a single-board computer, and would move to an Intel NUC at that point.
• Its onboard M.2 slot forces a choice between storage and wireless. There is no built-in Wi-Fi or Bluetooth, so adding them means using the M.2 slot for a combo card instead of an NVMe SSD; the Raspberry Pi 4 also keeps dual HDMI. The host found documentation and community answers scarce: the official Orange Pi OS folder was empty, and Armbian added a page only weeks after he got the board.
• Booting from NVMe works by flashing the bootloader to SPI flash. He boots Orange Pi's Ubuntu 22.04 desktop image from microSD, copies the image over by SCP, installs the bootloader to the 16 MB SPI flash through orangepi-config (System, Install, option 7), then writes the image to the NVMe with dd. Default login is root with password "orangepi"; he says the same process worked for Armbian.
• A reused SPI flash or microSD card will defeat the M.2 boot. SPI flash that already holds partitions throws a no-space error until they are deleted with gdisk, as must old partitions on the NVMe. The microSD's data volume shares the SSD's UUID and takes precedence if reinserted, so it needs a randomized UUID. The NVMe install is a fresh system with a new DHCP address and settings.
• The NVMe SSD benchmarked about ten times faster than microSD. On pibenchmarks.com his Kioxia 128 GB NVMe scored 19,924, about sixth among the site's fastest SSD averages (top: 24,307), against roughly 2,000 for SD cards. Cold boot from the SSD took about 10 seconds; the drive cost under $20 on Amazon, and the M.2 needed hot glue for lack of a fitting screw.

2022-12-30 · open the original ↗

• Browser internals are learnable, and you need not start with source. Lawrence, who says he has worked on browsers or web technology for about 25 years (IE, Fiddler, Chrome, Edge), points to bug trackers, design docs, blogs and documentation as ways in, and argues that reading C++ is much easier than writing it; Chromium's code is fairly straightforward to read, unlike IE's older network stack. He says he joined the IE team largely to see why Office's "clip of the day" would not change, printed the network stack's source, and found about six bugs, with about 100 other suspect spots circled.
• The decade's main browser fight is over becoming a full app platform. Lawrence says one camp wants the web to match native capability, while security and privacy are why the web succeeded; his view is that it has to be both, or the web stays a second-class citizen. He cites the rule that iOS browsers must use WebKit, which he says has shortcomings against Blink and Gecko, as a restriction the European Union is now questioning.
• IE's complexity lay in its compatibility ecosystem, not just its code. He says IE carried decades-old hacks (one for a CompuServe 1.0 beta) that nobody dared remove in case a line-of-business app relied on them. The day after he returned to Microsoft from the Chrome team he was asked whether to re-platform on Chromium, and he backed it; he puts a modern browser at about 12 million lines of C++ plus about as much test code.
• Chromium's layering keeps work simple, except for cross-cutting changes. Lawrence describes "content" (the base for building a multi-process browser), "Chrome" (the browser UI around it) and optional "components" such as the spell checker; Edge took the Chrome layer too, because Microsoft had no Win32 browser front end since IE. His hardest change carried one Boolean (whether a user had edited a form field on an HTTP page) across about 48 files, and he credits the modularity with making Electron and clean testing with fakes and mocks possible.
• Search revenue funds browsers and explains many of their decisions. Citing public figures from Mozilla and others, Lawrence says one point of browser market share is worth roughly $100–200 million a year, and that Google reportedly pays Apple about $12 billion a year to be the iPhone's search engine. He infers Chrome is worth tens of billions to Google, which is why browsers carry huge teams and why each defaults to its payer's search engine.

2022-12-28 · open the original ↗

• Nova kept Hachyderm open rather than turn Twitter refugees away. The server went from about 700 users on November 3, 2022 to 30,000 a couple of weeks later; Nova says people were breaking up with a part of their identity and they "didn't have it in me to say no". They intend to keep sign-ups open until perhaps 200,000 users, closing only briefly, as during recent DDoS threats; a legal entity is being formed behind it.
• Upload bandwidth at home, not compute, forced the move off the rack. One day saw a terabyte of egress against a small-business line advertised at a gigabit down and 200 Mb up, so serving users saturated the upload; the instance is moving to Hetzner bare-metal "auction servers". Nova describes cascading failures: bad disks slowed media reads, the Sidekiq queues behind them filled, and memory filled after that.
• Mastodon's waste is unscheduled queues and unplanned database access. Nova finds the Sidekiq queues arbitrary and overlapping, like old cron jobs, and tuned by hand (threads times concurrency kept under the DB pool); the Rails code reads and writes whenever it likes, which they call the biggest opportunity for improvement. They chose not to run it on Kubernetes, despite a Helm chart others use successfully, so they understand the application before possibly rewriting parts.
• Followers lost in migration are stalled federation handshakes, not losses. Every instance must re-verify each follow with HTTP signatures, so a neglected server's queue may never finish; Hanselman's 850 stranded followers sat on mastodon.online and qoto.org, and Nova says there is no admin button to retry. They call this a consequence of federation and advise large accounts to join an instance aware of it or run their own.
• Seniors should give juniors room to fail safely. Nova credits skateboarding with teaching them to take calculated risks and recover, which is why exposing a home server to the internet felt manageable, and wants that freedom for early-career people provided they work like scientists and write things down. Hanselman's own gloss is that psychological safety means not "flipping the bozo bit" on someone who asks a question.

2022-12-23 · open the original ↗

• The attacker took LastPass's backup copies of customer vault data. As reported, a breach in August yielded source code and technical data, which the hacker used to compromise an employee and obtain credentials and keys for a cloud backup storage service, separate from LastPass's main network, holding backups of company and user data.
• Stored passwords are encrypted, but weak master passwords can be cracked. LastPass says vault usernames, passwords, secure notes and form-fill data are AES-256 encrypted and need the master password, which it never knows; website URLs were unencrypted. The host warns weak master passwords could be brute-forced, expects phishing aimed at master passwords, and tells viewers not to believe LastPass's claim that strong ones face no danger.
• Changing the master password alone does not protect stored accounts. The host says to set a strong master password or passphrase and enable two-factor authentication, but then to change every password in the vault, most sensitive first (email, banking, cryptocurrency), and turn on two-factor authentication on each site wherever possible.
• The host's view is that users should reconsider LastPass altogether. He calls the breach entirely unacceptable for a company holding that level of responsibility over secure data, and asks viewers for alternatives rather than naming one.

2022-12-21 · open the original ↗

• Wi-Fi 6 improves crowded networks, not single-device speed or range. The host calls per-device speed gains the big misconception about 802.11ax: a fully Wi-Fi 6 device with a decent antenna may gain a little, but the standard's features squeeze more total throughput from the airwaves as device count rises. It uses both 2.4 GHz and 5 GHz; Wi-Fi 6E adds the newly opened 6 GHz band.
• Several features add capacity by sharing the channel more efficiently. Multi-user MIMO goes from 4x4 in Wi-Fi 5 to 8x8; OFDMA splits bandwidth into subcarriers so devices transmit simultaneously with variable-size packets; QAM rises from 256 to 1024 symbols. A new 160 MHz channel width raises data rates but is more interference-prone, so the host says to measure the local spectrum before choosing it.
• Other features cut interference and save client battery. BSS colouring lets nearby access points share a channel unless they share a colour; enhanced beamforming targets individual devices more precisely, raising beams from four to eight; target wait time and the uplink resource scheduler tell devices when to transmit and sleep, which matters most for phones and IoT devices.
• The host says most existing Wi-Fi 5 networks need not upgrade. Home users rarely have the density to notice, and he runs a mix of Wi-Fi 5 and 6 access points himself without issue; a business with around a hundred devices sees little benefit either. Event centres, theatres, churches and stadiums with hundreds of people do benefit, and anyone buying new should buy Wi-Fi 6.
• The host expects the next standard to be years from wide use. He says vendors already tout "Wi-Fi 7" but it is not formally recognized, that the next version is expected to be "Wi-Fi 8", 802.11be, with no Wi-Fi Alliance timeline, and notes Wi-Fi 6 arrived in 2019 but access points took a year or two to spread. He credits OpenAI's ChatGPT with helping research the video.

2022-12-15 · open the original ↗

• A portable paid hotspot could earn thousands at a disconnected event. The idea came from a three-day festival with no service where one booth sold internet at $20 for 10 minutes. His explicitly made-up numbers for a 10,000-person, three-day event (1% a day buying a 4-hour plan, 0.5% a 24-hour plan) give just under $3,300. He notes Starlink's terms explicitly prohibit resale and says he will not resell his own; LTE SIMs need their terms checked too.
• The host recommends Starlink for RVs, with LTE or point-to-point as fallbacks. The RV service is portable, has no data caps and can be paused, at $135 a month for months used. Alternatives are an LTE router such as the Pepwave MAX BR1, which he calls one of "our most popular choices", with omni or directional antennas, or a line-of-sight link from a nearby neighbour's connection.
• The whole kit runs from one PoE switch on a battery. A UniFi USW Lite 8 PoE powers an EdgeRouter X (via an 802.3af-to-24 V passive PoE converter), a Cloud Key Gen2 Plus for Protect, a G3 Flex camera as a theft deterrent and a UAP AC Mesh, at about 25 W. Starlink adds about 50 W; an EcoFlow Delta 1300 ran everything just under eight hours, and he judges solar impractical, leaving a generator or venue power.
• UniFi's guest hotspot handles payment but its portal editor is unreliable. The EdgeRouter's WAN must be eth4 because eth0 takes PoE in; the guest VLAN 51 SSID uses a 10/5 Mbps profile, client isolation and Cisco FamilyShield DNS, which must be in the pre-authorization allow list. He set a 10-minute free trial, 4 hours for $5.99 and 24 hours for $9.95 through Stripe, and tested the cap. On UniFi Network 7.3.76, "apply changes" sometimes fails; Ubiquiti told him 7.4 will improve it.
• A UniFi Dream Router adds application-layer blocking at lower power. The EdgeRouter setup cannot block streaming or torrenting, which he fears could bring piracy notices on his connection. The UDR's traffic management blocks peer-to-peer and media streaming, its "family" content filter blocks VPN, explicit and malicious domains, and it draws about 8 W. Against it: forum reports of lock-ups, rare stock, and his advice to carry two.

2022-12-14 · open the original ↗

• Build reliability first, then performance on top of it. Rao starts with failure mode analysis and the business cost of each failure, derives availability targets (SLA, RTO, RPO) and compares them with what the cloud guarantees, such as 99.99%. He then builds high availability into every tier: paired servers across Azure availability zones (at least three in most regions, each with its own power, cooling and network), clustering, database replication such as SQL Server Always On or Oracle Data Guard, plus disaster recovery and continuous monitoring.
• Performance design starts from a capacity model of business requirements. Rao lists content types, user base and locations, and targets such as a one- to two-second load time, then sizes resources against them. For SAP he uses SAP's Quick Sizer, which turns concurrent users and data size into a platform-neutral benchmark that he uses to shop for matching VMs.
• Close the user-to-cloud gap with proximity and private links. He moves compute or content near users via regions, CDN and Redis caching, and says site-to-site or point-to-site VPN mainly helps security rather than speed. For bulk loads, such as one terabyte of SAP data in a four-hour cutover window, he uses ExpressRoute, with committed bandwidth from 1–2 Gbps up to 100 Gbps (10 Gbps common); he says 7 or 8 of 10 large performance-oriented enterprise workloads use it, and nearly all of SAP.
• Horizontal scaling suits massive loads, with cloud autoscaling doing the work. Rao prefers adding nodes over enlarging one machine, though he notes most people still scale SQL Server and Oracle vertically; Giard adds the caveat, which Rao accepts, that the code and database must handle parallel requests or sharding. He recommends VM scale sets with thresholds (for example CPU above 90%) and the built-in autoscaling of AKS, Container Instances and Container Apps.
• Storage design is king for business-critical workloads. Rao says that in about seven of ten performance escalations he is pulled into, the fix is in storage design. He weighs IOPS, throughput and above all latency: 15,000 or 60,000 IOPS is no use at 7–8 ms when a database log drive such as SAP HANA's needs sub-millisecond. His tools are Ultra Disk, Premium SSD v2, Azure NetApp Files' ultra tier, the free Write Accelerator, striping data files across disks, and proximity placement groups to keep application and database servers physically close.

2022-12-10 · open the original ↗

• Starlink draws most of the Network in a Box's power. The whole setup pulls about 70–75 W from an EcoFlow Delta, of which about 25 W is the boxed gear (USW Lite 8 PoE switch powering a Cloud Key Gen2, EdgeRouter X, a pole-mounted mesh access point and a camera) and the rest Starlink; the host estimates the battery keeps it online for about 10 to 12 hours.

2022-12-09 · open the original ↗

• A complete small network can run from one power cord. A USW Lite 8 PoE switch, drawing 24 to 26 W in total, powers a Cloud Key Gen2 Plus running Protect, a G3 Flex camera, a mesh access point, and an EdgeRouter X through an 802.3af-to-24 V passive PoE converter; only the WAN needs plugging in.

2022-12-09 · open the original ↗

• Leadership rests on influence, with or without formal authority. Humphreys, who leads a team of cross-solution architects in Microsoft's U.S. Global Partner Solutions organization, says that as an individual contributor managing the technical relationship with one global customer she had to align Microsoft's internal teams and the customer's teams on a shared goal, and create clarity on why each was being asked to act. She cites a panel of Microsoft distinguished engineers and technical fellows, one of whom called "leading with authority" a myth, and agrees: managers at every level still get things done by understanding what motivates people and what "give-gets" are on offer.
• Asking others for help is a leadership skill, not cheating. Early on she felt she was cheating when she posted customer questions in Microsoft's internal MCS architect communities; a Microsoft leadership program taught her that pulling people in gives them a chance to have impact and broadens what you can do. She frames admitting what you don't know as courage, since knowing everything is as impossible as being perfect.
• Technical people are motivated by recognition, learning and room to fail. Humphreys says she has only ever managed technical teams and treats that as a requirement when considering a role, because as a former technical person she knows the motive is often not money. Her first management job was a "triple whammy": first-time manager, at Microsoft, leading her former peers as account technology strategists, which changed the relationship because she now affected their pay, promotion and career; she learned to ask each person what motivates them.
• You must find and own your mentoring relationships yourself. She kept a network of peer mentors inside and outside her organization, and does "speed mentorship," asking someone with a strength she lacks to walk her through how they think; she says going in with a stated purpose is the mentee's responsibility, and Giard adds that he once waited for mentors to come to him. As a manager, she assigns new team members peer mentors who can show them "what good looks like."
• Learn leadership from people you recognize as good leaders. Asked where to start, she recommends the book Radical Candor, by Kim Malone Scott, for being clear about feedback while also encouraging people. Mostly she learns by approaching people she sees as great leaders, whether managers, distinguished engineers or early-career staff, and through Microsoft's internal leadership programs and communities.