<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title><![CDATA[All Cloud Hub Blogs]]></title><description><![CDATA[All Cloud Hub Blogs]]></description><link>https://allcloudhub.hashnode.dev</link><generator>RSS for Node</generator><lastBuildDate>Sat, 21 Nov 2026 14:38:50 GMT</lastBuildDate><atom:link href="https://allcloudhub.hashnode.dev/rss.xml" rel="self" type="application/rss+xml"/><language><![CDATA[en]]></language><ttl>60</ttl><item><title><![CDATA[OneDrive killed my file dates: what actually changes on a move, and how to verify it before you delete the source]]></title><description><![CDATA[A migration ticket that reads "our files lost their dates" is almost never about the files. It is about one of four different dates being overwritten in…]]></description><link>https://allcloudhub.hashnode.dev/onedrive-killed-my-file-dates-what-actually-changes-on-a-move-and-how-to-verify-it-before-you-delete-the-source</link><guid isPermaLink="true">https://allcloudhub.hashnode.dev/onedrive-killed-my-file-dates-what-actually-changes-on-a-move-and-how-to-verify-it-before-you-delete-the-source</guid><category><![CDATA[OneDrive]]></category><category><![CDATA[Powershell]]></category><category><![CDATA[filemanagement ]]></category><category><![CDATA[CloudStorage]]></category><category><![CDATA[data migration]]></category><dc:creator><![CDATA[All Cloud Hub]]></dc:creator><pubDate>Sat, 21 Nov 2026 13:08:21 GMT</pubDate><content:encoded><![CDATA[<p>A migration ticket that reads "our files lost their dates" is almost never about the files. It is about one of four different dates being overwritten in transit, and the person filing it usually cannot say which.</p>
<p>That ambiguity is expensive. If you "fix" the wrong date, you re-upload an archive for nothing. If you ignore the right one, your retention schedule becomes fiction.</p>
<p>This post separates the four dates, shows what each migration method does to them, and gives you a verification routine that is cheap enough to run before you delete anything.</p>
<h2>Four dates, four owners</h2>
<table>
<thead>
<tr>
<th>Date</th>
<th>Where it lives</th>
<th>Who controls it</th>
</tr>
</thead>
<tbody><tr>
<td><strong>Date created</strong></td>
<td>Local filesystem metadata</td>
<td>The operating system of the device the file first landed on</td>
</tr>
<tr>
<td><strong>Date modified</strong></td>
<td>Local filesystem metadata</td>
<td>The application that last saved the content</td>
</tr>
<tr>
<td><strong>Service created / modified</strong></td>
<td>The provider's item record</td>
<td>The cloud service, when the item was created there</td>
</tr>
<tr>
<td><strong>Document-internal dates</strong></td>
<td>Inside the file bytes (Word properties, PDF metadata, an issue date on page one)</td>
<td>Whatever produced the document</td>
</tr>
</tbody></table>
<p>Microsoft's Graph API is unusually honest about the first and third. A <code>driveItem</code> exposes <code>createdDateTime</code> and <code>lastModifiedDateTime</code>, which the docs describe as <strong>"the created and modified date and time as seen from the service."</strong> A separate <code>fileSystemInfo</code> facet carries the value <strong>"reported by the device's local file system"</strong>, and Microsoft gives the canonical example: a file created on the device on <strong>Monday</strong> but uploaded on <strong>Tuesday</strong> will show a service creation date of Tuesday, while <code>fileSystemInfo.createdDateTime</code> can hold Monday if the uploading client writes it.</p>
<p>That last clause is the whole story. The service date is authoritative for the cloud. The client-provided date is an assertion, and it is only as good as the client that made it.</p>
<h2>What each method actually does</h2>
<p><strong>Windows copy vs move.</strong> Move within a volume: created and modified both survive, because it is the same file object. Copy anywhere, or move across volumes: modified survives, created becomes the time of the copy, and folders are re-dated. This is why an archive restored from a backup drive shows the restore date on every folder.</p>
<p><strong>Browser download and upload.</strong> Every file arrives as a new object. The service stamps it, and you inherit all the consequences. For a cross-account move, Microsoft's own support page states that files moved from OneDrive for work or school to a personal OneDrive <strong>"are consider new, so these files don't retain metadata details such as 'Date created' or 'Modified By'"</strong> and shared files must be re-shared.</p>
<p><strong>OneDrive web "Move to" vs "Copy to".</strong> Microsoft documents that <strong>Move to</strong> carries the document's history to the destination, and the source item goes to the source site's recycle bin when the move completes. <strong>Copy to</strong> copies only the latest version, so the destination begins a fresh version history. Different operations, different fidelity and neither is a promise about <code>Date created</code> across tenants.</p>
<p><strong>Native "Import" from other clouds.</strong> OneDrive's import from Google Drive, Dropbox and similar copies content into a <code>Files - Imports</code> folder and keeps the source untouched. Microsoft documents that on a re-run, a modified source file is compared by modified date and overwrites the target only if newer. It also notes that sharing properties do not come across. So the mechanism is date-aware for change detection, which is not the same as preserving your timestamps.</p>
<p><strong>Timestamp-preserving local copy.</strong> Robocopy, built into Windows, is the tool that actually keeps creation dates across volumes:</p>
<pre><code class="language-plaintext">robocopy "C:\Users\You\OneDrive - OldAccount" "C:\Users\You\OneDrive - NewAccount" /E /COPY:DAT /DCOPY:T /XJ /R:2 /W:2 /MT:16 /LOG+:copylog.txt
</code></pre>
<p><code>/E</code> recurses including empty directories, <code>/COPY:DAT</code> carries data, attributes and timestamps, <code>/DCOPY:T</code> preserves <strong>folder</strong> timestamps, <code>/XJ</code> avoids following junctions, <code>/R:2 /W:2</code> keeps retries sane, <code>/MT:16</code> parallelises, and <code>/LOG+:copylog.txt</code> gives you an audit trail you can attach to the ticket.</p>
<p>It requires both accounts synced on the same Windows machine with room for both trees which rules it out for large, restricted or cross-platform migrations.</p>
<h2>A verification routine you can run in ten minutes</h2>
<p>Before deleting anything, prove three things: the file exists, its content is intact, and the date you care about survived.</p>
<pre><code class="language-powershell"># 1. Inventory the source with the date you consider authoritative.
Get-ChildItem "C:\Users\You\OneDrive - OldAccount" -File -Recurse |
  Select-Object FullName, Length, CreationTime, LastWriteTime |
  Sort-Object FullName |
  Export-Csv -NoTypeInformation -Encoding utf8 source.csv

# 2. Build the same inventory at the destination.
Get-ChildItem "C:\Users\You\OneDrive - NewAccount" -File -Recurse |
  Select-Object FullName, Length, CreationTime, LastWriteTime |
  Sort-Object FullName |
  Export-Csv -NoTypeInformation -Encoding utf8 dest.csv

# 3. Compare on relative path, size and modified time.
$src = Import-Csv source.csv | ForEach-Object {
  [pscustomobject]@{ Rel = $_.FullName.Replace('C:\Users\You\OneDrive - OldAccount','');
                     Size = [int64]$_.Length; Modified = $_.LastWriteTime }
}
$dst = Import-Csv dest.csv | ForEach-Object {
  [pscustomobject]@{ Rel = $_.FullName.Replace('C:\Users\You\OneDrive - NewAccount','');
                     Size = [int64]$_.Length; Modified = $_.LastWriteTime }
}
Compare-Object $src $dst -Property Rel,Size,Modified -PassThru |
  Where-Object SideIndicator -eq '&lt;=' |
  Format-Table -AutoSize   # anything here exists at the source only
</code></pre>
<p>Two rules for reading the result:</p>
<ul>
<li>Treat <strong>size-and-path</strong> as the completeness check and <strong>modified time</strong> as the ordering check. Do not compare <code>CreationTime</code> that is the value you expect to change, and flagging it will produce a wall of false positives.</li>
<li>Run the comparison against <strong>OneDrive on the web</strong>, not only the local folder. File Explorer reports the local copy; the web view reports the service. When they disagree, the service is what every other device and every other person sees.</li>
</ul>
<p>For a content-level check on a chosen sample, hash a handful of files at both ends:</p>
<pre><code class="language-powershell">Get-FileHash "C:\Users\You\OneDrive - NewAccount\2023\contract-acme.pdf" -Algorithm SHA256
</code></pre>
<p>A matching SHA-256 proves the bytes are identical. It proves nothing about dates, permissions, sharing links or version history those are separate objects that a copy does not recreate.</p>
<h2>When dates are already gone, do not fake them</h2>
<p>An overwritten <strong>Date created</strong> cannot be recovered; the original value was never stored anywhere else. Two responses are honest, and one is not.</p>
<pre><code class="language-powershell"># Honest repair: remove the misleading "created today" value.
Set-Location "C:\Users\You\OneDrive - NewAccount"
Get-ChildItem -File -Recurse |
  ForEach-Object { $_.CreationTime = $_.LastWriteTime }
</code></pre>
<p>That makes created and modified agree, so date sorting behaves predictably. It does not invent history.</p>
<p>For anything governed by a retention rule, stop depending on filesystem dates entirely: put the date in the filename (<code>2023-03-14_invoice_acme.pdf</code>), because filenames survive every copy, and the document-internal date survives too. The dishonest option backdating a file to look like it was created earlier is the one that turns a migration problem into an audit problem.</p>
<h2>Where a cloud-to-cloud move fits</h2>
<p>Everything above assumes you can get both trees onto one Windows machine. Often you cannot: the source is a work account with external sharing blocked, the destination is on a Mac, or a 60 GB archive does not fit on a laptop with 12 GB free.</p>
<p><a href="https://allcloudhub.com/connect-onedrive/">All Cloud Hub</a> covers that case: connect both OneDrive identities with an alias for each, then move folders between them inside one workspace no ZIP, no local staging, and no second Windows PC. When you check the result, its <a href="https://allcloudhub.com/why-cloud-file-transfers-fail-and-how-to-reduce-the-risk/">guide to partial transfers and why they fail</a> is the right reading: verify the destination files rather than trusting a progress bar, because a transfer that reports success can still have dropped a subfolder.</p>
<p>Use it for what it is good at: moving the archive without a download detour, verifying both accounts from one workspace, and moving one year at a time so a sample can be checked before the source copies are removed. Do not expect it or anything else to guarantee a preserved <strong>Date created</strong> across accounts. Verify your sample, keep the internal document date, and put the load-bearing date somewhere no platform can overwrite.</p>
<p>The ticket closes with evidence, not with a feeling: a comparison report, a hash of a sample, the name of the date you chose to trust, and a note saying where the originals still are.</p>
]]></content:encoded></item><item><title><![CDATA[Five real costs that hide behind a cheap cloud storage plan]]></title><description><![CDATA[Cloud storage is priced like a subscription and used like an archive. That mismatch is where the money goes. A team upgrade usually follows the same…]]></description><link>https://allcloudhub.hashnode.dev/five-real-costs-that-hide-behind-a-cheap-cloud-storage-plan</link><guid isPermaLink="true">https://allcloudhub.hashnode.dev/five-real-costs-that-hide-behind-a-cheap-cloud-storage-plan</guid><category><![CDATA[CheapestCloudStorage]]></category><category><![CDATA[CloudStorage]]></category><category><![CDATA[AllCloudHub]]></category><category><![CDATA[CloudStorageComparison]]></category><category><![CDATA[multicloud]]></category><category><![CDATA[Multi-cloud strategy]]></category><dc:creator><![CDATA[All Cloud Hub]]></dc:creator><pubDate>Fri, 20 Nov 2026 12:34:06 GMT</pubDate><content:encoded><![CDATA[<p>Cloud storage is priced like a subscription and used like an archive. That mismatch is where the money goes.</p>
<p>A team upgrade usually follows the same script. A drive fills, someone compares two plans, sees that the larger one costs less per gigabyte, and buys it. Three months later the same drive is full again and nobody can say where the space went because the space never went anywhere. It is sitting in a second account nobody checked, in a duplicate photo library, and in a completed client project that was never moved off an active drive.</p>
<p>This is written for the person who owns that problem: the freelancer with four clients' accounts, or the small team whose files are spread across Google Drive, OneDrive, Dropbox and a pCloud archive. Five pain points, what each one actually costs, and the setup that fixes them.</p>
<h2>Pain point 1: The subscriptions multiply, the visibility doesn't</h2>
<p>The typical account list: Google Drive for documents, iCloud for a phone backup, OneDrive because Microsoft 365 came with it, Dropbox because one client sends folders there, and a pCloud or MEGA account opened for a single large transfer and never closed.</p>
<p>Individually each is affordable. Together they are a monthly cost with no owner and no dashboard. Search is per provider, so “where is the signed contract?” has five possible answers and no single query.</p>
<p>The structural problem is that <strong>free allowances never pool</strong>. Google's 15 GB, Apple's 5 GB and Microsoft's 5 GB do not become 25 GB of usable space; they remain separate quotas, and the first one to fill forces an upgrade while the others sit idle.</p>
<p><strong>What it costs:</strong> capacity you already rent, plus an hour a week of hunting for files you know you have.</p>
<h2>Pain point 2: You budget for signup pricing and renew at the real price</h2>
<p>Every comparison quotes the entry price. The invoice arrives at the renewal price, a year later, when nobody is watching.</p>
<p>Two mechanisms to price separately:</p>
<ul>
<li><strong>Term discount.</strong> Microsoft 365 Basic in India is ₹1,499 annually or ₹149 monthly. The annual plan saves ₹289 a genuine saving only if the full twelve months are used.</li>
<li><strong>One-time plans.</strong> pCloud currently lists a Black Friday lifetime price of $379 for 2 TB against a $709 reference price, with the annual plan at $99.99. Break-even: about <strong>3.8 years</strong> at the promotional price, <strong>7.1 years</strong> at reference. pCloud defines “lifetime” as 99 years or the account holder's lifetime, whichever is shorter.</li>
</ul>
<p><strong>What it costs:</strong> the difference between the number you approved and the number that renewed.</p>
<h2>Pain point 3: You buy headroom for work that is already finished</h2>
<p>Same provider, two tiers: 200 GB at ₹219 a month adds 195 GB after the free tier about <strong>₹1.12</strong> per usable gigabyte. 2 TB at ₹749 works out to roughly <strong>₹0.37</strong> per gigabyte.</p>
<p>The 2 TB plan is three times better value per gigabyte and <strong>₹6,360 more expensive annually</strong>. Both facts are true. What decides which one is “cheap” is the composition of your storage, and in most audits it looks like this: 140 GB of delivered RAW files, 60 GB of duplicate phone photos, 20 GB of exports of exports, and the actual active work sitting in a folder under one gigabyte.</p>
<p>Paying to keep delivered work available is a reasonable archiving decision. It is not a capacity decision, and the two have different price tags.</p>
<p><strong>What it costs:</strong> the annual difference between a plan that matches your working set and one sized for your archive.</p>
<h2>Pain point 4: Exit costs more than entry</h2>
<p>Nobody prices migration when they compare plans, and that is where the weekend goes.</p>
<p>Move a 300 GB project library the manual way download to a laptop, upload to the destination and you are moving 2,400,000 megabits each direction. On a steady 20 Mbps connection, about <strong>33 hours</strong>; at 50 Mbps, about <strong>13 hours</strong>. Both assume the line is perfect and nobody is streaming. Interrupt one folder and the destination looks complete while a subfolder is missing.</p>
<p>Cancelling a plan has a schedule too. pCloud: five days' grace, downgrade to free, over-quota files kept 30 days, then moved to Trash for another 30. Sixty-five days between “I cancelled” and “the files are gone”, which is exactly long enough to stop thinking about it.</p>
<p><strong>What it costs:</strong> evenings, bandwidth, and the risk of finding a gap after the recovery window.</p>
<h2>Pain point 5: The meters that are not storage</h2>
<p>The plan page sells capacity. Three other things can bill you.</p>
<ul>
<li><strong>Shared-link traffic.</strong> pCloud's 2 TB plan includes 4 TB of link traffic a month. Over the limit, links keep working but free pCloud recipients cannot download until the allowance resets or you buy more. A client-gallery workflow can hit this with the drive half empty.</li>
<li><strong>Retention add-ons.</strong> Extended File History in pCloud stretches recovery to 365 days, but only from activation; it cannot recover anything deleted more than 30 days before you bought it. Bought after an incident, it recovers nothing.</li>
<li><strong>Currency and tax.</strong> $99.99 is not ₹99.99 or €99.99; card conversion and local tax are part of what you paid.</li>
</ul>
<p><strong>What it costs:</strong> a bill for a feature you assumed was included.</p>
<h2>What actually fixes it: one workspace over the accounts you already rent</h2>
<p>Every one of those pain points traces back to the same gap no view across accounts, so the only lever anyone reaches for is “buy more capacity”.</p>
<p><a href="https://allcloudhub.com/pricing/">All Cloud Hub</a> closes that gap. It is a single workspace over cloud accounts you already own: Google Drive, Google Shared Drive, Google Photos, OneDrive, Dropbox, pCloud, MEGA, Backblaze B2, iCloud Drive and iCloud Photos. Files are searched, copied, moved and scheduled between providers from one dashboard, without routing them through a laptop.</p>
<p>Mapped against the five pain points:</p>
<table>
<thead>
<tr>
<th>Pain point</th>
<th>What the workspace gives you</th>
</tr>
</thead>
<tbody><tr>
<td>Five accounts, five search boxes</td>
<td>One search across every connected drive, with filters for account, type, date and tags; virtual folders group a project without copying the files.</td>
</tr>
<tr>
<td>Storing 400 GB while renting 3 TB</td>
<td>See actual usage per provider next to the others, so an upgrade decision is made with four numbers instead of one red bar.</td>
</tr>
<tr>
<td>Finished work squatting in an active drive</td>
<td>Copy or move completed projects to the account with room, including between two accounts on the same provider.</td>
</tr>
<tr>
<td>A migration that would take 33 laptop hours</td>
<td>Transfers run cloud-to-cloud between providers; a 300 GB archive does not consume your connection end to end.</td>
</tr>
<tr>
<td>Afraid to cancel a plan</td>
<td>Move what matters to the keeper account, verify it opens, then cancel on your own timeline instead of renewing from fear.</td>
</tr>
</tbody></table>
<p>Two honest limits, because they matter more than the feature list. <strong>A copy is a copy:</strong> it frees no space at the source until you remove the original, and the destination needs room before you start. <strong>Support varies by platform:</strong> iCloud Drive and iCloud Photos connect on web and iOS but not the native Android app, and Google Photos supports copy rather than move.</p>
<h2>What it costs, and how to start</h2>
<p>The Free plan is permanent: <strong>3 cloud accounts</strong>, up to <strong>5 GB of transfers a month</strong>, manual folder sync. Enough to prove the idea on your own files rather than a demo. Power User is <strong>$19/month or $49/year</strong> with unlimited connected accounts, <strong>100 GB of transfers a month</strong>, automatic folder sync, faster transfers and priority support. Scheduled transfers are available on the Free plan too.</p>
<p>Start with the free tier and a small, real task: move one finished client project into the account you intend to keep, verify it opens at the destination, then delete the source. If that feels safe, scale it to the archive.</p>
<p>When the move itself is the worry, All Cloud Hub's <a href="https://allcloudhub.com/how-to-transfer-files-between-cloud-storage/">cloud-to-cloud transfer guide</a> covers what happens during a transfer, Copy versus Move, and how to verify the destination before removing anything.</p>
<h2>The audit that replaces the upgrade</h2>
<ol>
<li><strong>Inventory</strong> every subscription, its renewal date and its amount.</li>
<li><strong>Connect</strong> the accounts and compare real usage side by side.</li>
<li><strong>Identify</strong> the overlap: duplicated libraries, dual-synced project folders, work in the wrong account.</li>
<li><strong>Move</strong> finished work to the keeper account and verify at the destination.</li>
<li><strong>Cancel</strong> what is no longer load-bearing, after confirming nothing lives only there.</li>
<li><strong>Re-decide</strong> whether you need a bigger plan. Frequently the answer is no.</li>
</ol>
<p>Cheap cloud storage is not the lowest price per gigabyte. It is the arrangement where nothing you already pay for is idle, no subscription renews unexamined, and leaving a provider costs an evening instead of a weekend.</p>
]]></content:encoded></item><item><title><![CDATA[Debug a full OneDrive with one Microsoft Graph quota request]]></title><description><![CDATA[A support ticket says, “The archive copy finished, I deleted the source, and OneDrive still won’t accept a 250 MB upload.” Those are three observations…]]></description><link>https://allcloudhub.hashnode.dev/debug-a-full-onedrive-with-one-microsoft-graph-quota-request</link><guid isPermaLink="true">https://allcloudhub.hashnode.dev/debug-a-full-onedrive-with-one-microsoft-graph-quota-request</guid><category><![CDATA[MicrosoftGraph]]></category><category><![CDATA[OneDrive]]></category><category><![CDATA[Python]]></category><category><![CDATA[CloudStorage]]></category><category><![CDATA[Microsoft365]]></category><category><![CDATA[Cloud Storage Management]]></category><category><![CDATA[multi-cloud]]></category><dc:creator><![CDATA[All Cloud Hub]]></dc:creator><pubDate>Thu, 19 Nov 2026 05:15:17 GMT</pubDate><content:encoded><![CDATA[<p>A support ticket says, “The archive copy finished, I deleted the source, and OneDrive still won’t accept a 250 MB upload.” Those are three observations, not proof that 250 MB became available. A copy job can succeed without reclaiming a byte. A deletion can reach the online Recycle bin without reducing usage. And a quota report can lag behind both events.</p>
<p>Here’s a diagnostic for an <strong>existing OneDrive you’re authorized to inspect</strong>. It reads the quota, not the file contents. There’s no delete endpoint in the workflow.</p>
<h2>Read the right drive before listing any files</h2>
<p>First check the affected account in OneDrive on the web. Is the source still in <strong>My files</strong>, in the <strong>online Recycle bin</strong>, or in neither? If Windows <strong>Free up space</strong> made the file online-only, it never left the cloud. If the source was removed from a synced folder while offline, check whether that change has reached the website. Neither case calls for a second deletion.</p>
<p>In <a href="https://developer.microsoft.com/en-us/graph/graph-explorer">Microsoft Graph Explorer</a>, sign in to the affected account rather than using the sample tenant. Set the method to <strong>GET</strong>, paste the URL below into the address field, and run it with no request body:</p>
<pre><code class="language-http">GET https://graph.microsoft.com/v1.0/me/drive?$select=driveType,quota
</code></pre>
<p>Use delegated <code>Files.Read</code>, the least-privileged permission listed for this <a href="https://learn.microsoft.com/en-us/graph/api/drive-get?view=graph-rest-1.0">Get drive request</a>. It doesn’t need write or tenant-wide permissions. If your organization blocks Explorer, use its approved tooling or the web storage view. <code>/me/drive</code> is the signed-in user’s default drive; it isn’t a reading for a separate team’s SharePoint library. Use an existing drive: Microsoft notes that this call can provision a licensed user’s drive if it doesn’t already exist.</p>
<p>Here’s a <strong>synthetic response</strong> for a personal account. These values are bytes, not a screenshot or measurement from a customer:</p>
<pre><code class="language-json">{
  "driveType": "personal",
  "quota": {
    "total": 5000000000,
    "used": 5200000000,
    "remaining": 0,
    "deleted": 350000000,
    "state": "exceeded"
  }
}
</code></pre>
<p>The crucial field is <code>deleted</code>. It reports bytes consumed by files <strong>currently in the recycle bin</strong>, not a running total of bytes you’ve reclaimed. Don’t add it to <code>used</code>. For OneDrive Personal, usage reflects Microsoft’s unified cloud quota across services, so summing visible files will not necessarily reproduce it. The <a href="https://learn.microsoft.com/en-us/graph/api/resources/quota?view=graph-rest-1.0">quota resource reference</a> defines each field.</p>
<p>A transfer dashboard can use bytes to measure something else entirely. For example, All Cloud Hub’s <a href="https://allcloudhub.com/understand-bandwidth-storage-usage/">usage reference</a> distinguishes bytes transferred from bytes stored. “350 MB copied” is not evidence that Microsoft reclaimed 350 MB.</p>
<h2>Calculate the shortfall, not just the free space</h2>
<p>A <code>remaining</code> value of zero doesn’t tell you how far over quota you are. Save the sample above as <code>quota.json</code>, or use your own response. Save this script as <code>quota_check.py</code> beside it and run <code>python quota_check.py</code> with Python 3:</p>
<pre><code class="language-python">import json
from pathlib import Path

data = json.loads(Path("quota.json").read_text(encoding="utf-8"))
q = data.get("quota") or {}
if q.get("used") is None or q.get("total") is None:
    raise SystemExit("Quota unavailable. Check OneDrive on the web.")

upload = 250_000_000  # 250 decimal MB
needed = max(0, q["used"] + upload - q["total"])
print(f"Bytes to reclaim for this upload: {needed:,}")
bin_bytes = q.get("deleted")
reported = "unknown" if bin_bytes is None else f"{bin_bytes:,}"
print(f"Bytes reported in bin: {reported}")
</code></pre>
<p>For this sample, the output is <strong>450,000,000 bytes to reclaim; 350,000,000 bytes in the bin</strong>. The account starts 200 MB over. Even if all 350 MB were disposable and stopped counting, only 150 MB would be free: the upload would still be 100 MB short. The arithmetic is a lower bound assuming no new writes and an unchanged limit, <strong>not</strong> a recommendation to purge the bin.</p>
<h2>Choose the next check from the evidence</h2>
<ul>
<li><strong>Personal account, item still in the bin:</strong> Restore anything questionable. For a confirmed disposable item, select <em>that item</em> in the online bin and choose Delete. Record the time. Microsoft says personal quota updates after cleanup can take <strong>up to 24 hours</strong>; don’t interpret an immediate second reading as a failed delete.</li>
<li><strong>Personal account, small bin, high usage:</strong> Check the Microsoft storage breakdown for <a href="http://Outlook.com">Outlook.com</a> attachments, and compare the current allowance with the user’s subscription. Missing <code>deleted</code> data means <em>unknown</em>, not zero.</li>
<li><strong>Work or school account, empty user bin, high usage:</strong> Give IT the affected drive or site, quota figures, and timestamps. Ask which limit was hit and whether retention applies. A personal account’s update window is not a service guarantee for a company site.</li>
</ul>
<p>Two enterprise details can save a needless purge. For ordinary files in SharePoint Online, <a href="https://support.microsoft.com/en-us/sharepoint/admin/restore-deleted-items-from-the-site-collection-recycle-bin">Microsoft distinguishes two bins</a>: <strong>the first stage counts toward the site’s storage-used metric; the second does not</strong>. The usual 93-day recovery period runs from the original deletion across both stages. It is not a 93-day wait for a quota update. Separately, a policy can keep a copy in a <a href="https://learn.microsoft.com/en-us/purview/retention">Preservation Hold library</a>, which <em>does</em> count against the site quota. Keep the hold intact and have the administrator confirm what accounts for the space.</p>
<p>If the archive was a scheduled job, inspect both the operation and the result. All Cloud Hub’s <a href="https://allcloudhub.com/view-transfer-run-history/">Run History guide</a>, for example, distinguishes <strong>Partial</strong> from <strong>Succeeded</strong>. A successful <strong>Copy</strong> leaves the source in place; a partial copy is a reason to keep the source. Open the archive at its destination before anyone retires the original.</p>
<p>Close the ticket with the account type, time of each reading, <code>used</code> and <code>total</code> before and after, and the space needed for the blocked upload. Keep access tokens and private filenames out of the ticket. “The archive ran” describes a job; “the account now has enough room for the upload” answers the user’s question.</p>
]]></content:encoded></item><item><title><![CDATA[Free Google Drive Space With a Project Archive You Can Actually Restore]]></title><description><![CDATA[There’s a folder named Archive. The filenames look right. The transfer job finished. Now support needs last quarter’s release bundle: can anyone actually…]]></description><link>https://allcloudhub.hashnode.dev/free-google-drive-space-with-a-project-archive-you-can-actually-restore</link><guid isPermaLink="true">https://allcloudhub.hashnode.dev/free-google-drive-space-with-a-project-archive-you-can-actually-restore</guid><category><![CDATA[Google Drive]]></category><category><![CDATA[Cloud Storage Management]]></category><category><![CDATA[multi-cloud]]></category><category><![CDATA[ cloud management]]></category><dc:creator><![CDATA[All Cloud Hub]]></dc:creator><pubDate>Wed, 18 Nov 2026 09:45:51 GMT</pubDate><content:encoded><![CDATA[<p>There’s a folder named <code>Archive</code>. The filenames look right. The transfer job finished. Now support needs last quarter’s release bundle: can anyone actually get it back after the original leaves Google Drive?</p>
<p>Use a stronger test: <strong>could someone restore one named artifact without using your Drive account or asking you for access?</strong> If the answer is yes, you have a candidate for freeing space. If the answer is no, keep the source. Here’s a release-close runbook you can reuse.</p>
<h2>1. Set a target smaller than “clean up Drive”</h2>
<p>Start in Google One’s storage breakdown. If Gmail or Photos accounts for most of the usage, archiving a release folder may be the wrong job. If Drive is the issue, open <strong>Drive → Storage</strong> and identify large files you own. A Shared Drive managed by your employer is a different case: follow the organization’s storage, retention, and access policies, and use an approved archive destination.</p>
<p>Suppose the next build needs 2 GB. Here’s an illustrative release folder and the decision for each item:</p>
<table>
<thead>
<tr>
<th>Item</th>
<th>What someone will need later</th>
<th>Decision</th>
</tr>
</thead>
<tbody><tr>
<td><code>release-bundle.zip</code> · 1.5 GB</td>
<td>The exact binary that shipped</td>
<td>Archive it and compare hashes.</td>
</tr>
<tr>
<td><code>demo-master.mp4</code> · 1.2 GB</td>
<td>A playable recording, including the final demo</td>
<td>Archive it and test playback.</td>
</tr>
<tr>
<td>Support handoff · Google Doc</td>
<td>The URL in the support handbook, comments, and edit history</td>
<td>Keep the live original in Drive.</td>
</tr>
</tbody></table>
<p>The two binaries are candidates for roughly <strong>2.7 GB</strong> of relief. That’s enough for this example; there’s no reason to move the Google Doc. An export that looks right is still not the collaborative document the support handbook links to.</p>
<h2>2. Copy to somewhere that won’t mirror a deletion</h2>
<p>Use a second cloud account with enough room, or an external disk outside your Drive sync folders. Don’t use a two-way mirror as the sole archive: retiring the source could propagate the deletion. For irreplaceable release artifacts, arrange a second independent backup as well.</p>
<p>For a cloud destination, one option is <strong>All Cloud Hub</strong>, a file manager for existing cloud accounts. Its <a href="https://allcloudhub.com/how-to-transfer-files-between-cloud-storage/">cloud-to-cloud transfer guide</a> describes two different operations: <strong>Copy</strong> leaves the Drive original in place; <strong>Move</strong> removes it after transfer. Choose <strong>Copy</strong> here. The destination needs enough storage, and the batch must fit the tool’s remaining transfer allowance. A manual download and upload works too.</p>
<p>Record the destination account, folder, filenames, and expected sizes in the release ticket. Don’t mark the ticket done yet. Copying 2.7 GB into another account initially frees <strong>0 GB</strong> in Google Drive.</p>
<h2>3. Restore the artifact, not the shortcut</h2>
<p>Open the destination in its own app. Download <code>release-bundle.zip</code> into a fresh local directory called <code>restored/</code>. Compare it with the trusted build output in <code>original/</code>, not with a second download of the archive. If your release process already recorded a trusted SHA-256 checksum, use that instead.</p>
<p>On Linux or WSL with GNU coreutils, run this from the directory containing <code>original/</code> and <code>restored/</code>:</p>
<pre><code class="language-bash">(cd original &amp;&amp; sha256sum release-bundle.zip) &gt; bundle.sha256 &amp;&amp;
  (cd restored &amp;&amp; sha256sum -c ../bundle.sha256)
</code></pre>
<p>On a byte-for-byte match, the check prints <code>release-bundle.zip: OK</code>. On a mismatch or missing file, stop and keep the Drive source. A match verifies the bytes, not their usefulness: extract the ZIP and check that the expected artifact is inside. For the video, play the beginning and the final minute with sound. For the handoff Doc, test the <strong>existing</strong> link from a teammate’s account. Those are three different acceptance tests.</p>
<p>If this was a scheduled copy, inspect the execution record too. All Cloud Hub’s <a href="https://allcloudhub.com/view-transfer-run-history/">Run History guide</a> distinguishes <strong>Partial</strong> from <strong>Succeeded</strong>. Seventeen copied files out of eighteen is not permission to delete an eighteen-file source folder.</p>
<h2>4. Retire the source and leave directions</h2>
<p>After verification and any required sign-off, remove only the archived binaries from Drive. Keep the live Doc and any item that failed a check. In Drive, trashed files still use quota until permanently deleted. If you need the space immediately, select just the verified items in <strong>Trash</strong> and delete those permanently. If you can wait, Google normally clears Drive Trash after 30 days. A large cleanup can also take time to show up on the storage meter. <a href="https://support.google.com/drive/answer/2375102?hl=en">Google’s deletion guide</a> explains what happens to the files and to other people’s access.</p>
<p>Close the ticket with four details: <strong>where the archive lives, who can retrieve it, what was restored successfully, and how much Google storage became available</strong>. If the next maintainer can recover the release from that note, you’ve freed space without making yourself part of the backup system.</p>
]]></content:encoded></item></channel></rss>