UltaHost Ruined My Business and Cost Me Thousands of Dollars: Never Choose This Service

Incident record

On 2 October 2026, I asked UltaHost to move my server onto an official cPanel licence. The reinstall ran on 5 October. I expected a routine job with a verified backup in hand. Instead, the cPanel restore finished without my main database, the site was offline for most of a day, and the server was later suspended over an invoice I say I never received. This page covers the parts of the support ticket that matter, with the screenshots, so readers can judge it for themselves. All times below are as displayed on the UltaHost ticket pages; where a reply sits above a later message in a screenshot, the figure caption says so.

Background: what I had and what I asked for

I bought UltaHost's POWER Ultra plan for this server. The ticket lists it as the service on the account.

vynex.ai is a live SaaS. Customers create accounts, buy credits, and generate videos and images from the site. The application and its MySQL/MariaDB database live on a single UltaHost server that runs WHM and cPanel. The server hosted two cPanel accounts. One of them, xyz, held the whole application, including the database xyz_vynex with every customer account and every order.

The server already had an installed copy of WHM/cPanel, but it was not running on an official licence. I wanted to fix that, and I asked UltaHost to explain the procedure, the cost, the downtime and whether my data could be migrated.

That was a reasonable request. The mistake was not the request. The mistake was what happened after it.

2 October: the request and the price list

Phase 1

I opened ticket #2026100200275, titled "Switching to Official cPanel", at 9:00 AM on 2 October. I asked for the full procedure, the total cost, the downtime, and whether my existing data would be migrated or whether I had to back it up and restore it myself.

Support replied with a price list for the WHM/cPanel licence (monthly):

• cPanel Solo (1 account): $15
• cPanel Admin (5 accounts): $25
• cPanel Pro Cloud (30 accounts): $36
• cPanel Premier Cloud (100 accounts): $56

I asked for a fresh cPanel install rather than a licence on the existing one, because I wanted to be sure the installation was clean. Support then said that since I had two accounts on the server, I should buy the Admin licence and that they could back up the data first and restore it after the reinstall.

That was my reply that a Solo licence was enough for me. The ticket does not show support confirming a Solo licence after that.

"I don't need 2 accounts solo account is good for me. How much time will it take to install new cPanel backup/restore?"

That was my reply that a Solo licence was enough for me. The ticket does not show support confirming a Solo licence after that.

4 to 5 October: the backup, the maintenance mode, the payment

Phase 2

On 4 October I cleaned up files on the server to shrink the cPanel account. By my message at 9:11 PM it was about 46 GB, and I asked when the install could happen. Support said the xyz account was using about 135 GB and that they could not give an ETA because it depended on data size, location and traffic. They offered to do it during off-peak hours.

Later on 4 October, support created a backup of the xyz account and sent me the location of the backup file. They asked me to confirm that I had a full backup of my own, because the reinstall would remove everything on the server.

On 5 October support asked me to put the website into maintenance mode (2:36 PM), and warned that the server data would be removed during the reinstall. At 2:57 PM I wrote that two users were still active and I would switch to maintenance once they finished. At 3:01 PM I wrote that the app was in maintenance and asked them to start as soon as possible.

At 5:59 PM a billing agent told me an additional invoice had been generated for the cPanel licence, and that "the backup cannot be restored without an active cPanel license." They asked me to keep the site offline until the payment was made.

I paid that invoice at 6:05 PM and asked for the restore and activation to be done straight away.

5 to 6 October: the restore that "completed" without my database

Phase 3

After payment, the ticket shows a long gap. At 6:19 PM I asked for an estimate, and at 6:27 PM I was told the request had been processed. At 8:52 PM I asked again for the restore status, wrote that the server had been out for about five hours with no progress update, and asked for escalation to a senior technician or management.

At 9:43 PM I wrote that the business had been offline for about six hours and listed the facts: the licence invoice was paid, the team had said the backup would be restored once the licence was active, and since then I had only received generic replies.

At 10:09 PM a technical administrator said they were working on the concern. At 10:15 PM I wrote that it had been almost seven hours of downtime. At 10:45 PM the same administrator said the cPanel reinstall and backup were being moved and that they expected it to finish within 1.5 to 2 hours.

At 10:52 PM I asked why the restore had not started for five hours after I paid at 6:05 PM, and why it needed another two hours. The support manager again replied that the inquiry was under review.

At 11:41 PM the technical administrator reported that the restore was complete: "The cPanel account, along with all data present in the supplied backup, has been restored to the server. Web and database services are operational, and most of the hosted websites are now responding normally. The original cPanel license is also active on the server."

I replied at 11:59 PM thanking them and saying I would check all websites and services and report any issues.

That reply was premature. The restored backup did not contain my production database.

6 October: the September 30 backup and the long morning

Phase 4

Shortly after midnight on 6 October, I wrote a detailed message to the technical administrator. The reinstall had been done from the cPanel account backup, and that backup did not include the database xyz_vynex. I explained that this database existed on the old installation under MariaDB but was not mapped to the cPanel account, so the account backup skipped it. It held every customer account and order for vynex.ai. I asked for three things: a full-disk snapshot or backup from before the reinstall, the old MySQL data directory and binary logs if they still existed, and an answer on whether the old disk had been wiped, reformatted or was still a cloud volume.

At 12:37 AM I sent a message marked "Urgent: recover a MySQL/MariaDB database from the old disk". A support manager replied at 12:57 AM that they were reviewing the matter. At 1:03 AM I wrote that I had raised the issue before 4 PM the previous day, that the old disk had been wiped without a full server image being taken first, and that the site had been down since the reinstall.

At 1:33 AM the technical administrator said they were checking and asked me to hold. At 1:51 AM I wrote that more than nine hours had passed since I first raised the issue.

At 2:36 AM UltaHost reported what they had found:

"The latest full-disk backup available is dated September 30, 2026. The backup that was restored during the recent cPanel restoration contains the data that was present in that cPanel backup, but the xyz_vynex database you are referring to is not present in the latest restored backup."

At 6:19 AM I wrote that I had more than 300 registrations in the previous four or five days and asked them to restore xyz_vynex from the 30 September backup. At 6:47 AM a support agent replied:

"The cPanel account backup did not include xyz_vynex, and we should have verified that it was backed up before reinstalling. As requested, we will check the September 30, 2026 full-disk backup and restore only xyz_vynex if it is present."

At 6:52 AM I set out my demands: restore the database from the 30 September backup immediately; check every other possible source for data after 30 September; send a written incident report; and pay full compensation including a refund of the extra cPanel licence invoice. I also wrote that the downtime was around 20 hours at that point.

The next hour went round in circles. At 6:59 AM support said their earlier message was "an initial acknowledgment" and the investigation was still in progress. At 7:03 AM I asked for a yes or no on whether xyz_vynex was in the backup. At 7:12 AM support said they had not yet confirmed that, and that they could not give a restore time until recovery availability was confirmed. At 7:19 AM I replied that a check of one backup for one database should take minutes, not hours. At 7:27 AM a new agent said they had only just taken over the ticket.

At 7:36 AM the agent wrote: "We found xyz_vynex in the September 30 backup. Please see the attached screenshot." They offered two options: restore the backup on a separate server and give me SSH access to verify and extract the database, or extract it and copy it to my current server. I chose the first option at 7:41 AM and sent a public key so that I could export the database myself.

6 October: the temporary server and my own recovery

Phase 5

At 8:09 AM I asked how long the restore to the temporary server would take. I wrote that the site had been down for over 21 hours. At 8:30 AM support said the secondary temporary server was enabled and sent a link to the root details and a screenshot of the database. At 8:32 AM they said the temporary server would remain available for seven days.

At 3:13 PM I sent a formal complaint to UltaHost management. I wrote that I had recovered the database myself from the 30 September disk image, and had then spent another day fixing the rebuilt server, with the cron jobs disabled, the wrong PHP version in use, mail broken and the admin tools missing. I listed the impact I was claiming and asked for a management-level reply within three business days.

6 to 8 October: the compensation offers

Phase 6

At 8:17 PM on 6 October the support manager offered two weeks of free hosting as compensation. At 5:40 PM the same day, a different agent had written that the team was working on the issue.

At 8:48 AM on 7 October I rejected the offer. I listed the losses as I saw them: a traffic drop of roughly half in a single day, six days of customer data gone, around 20 hours of downtime and recovery work, and damage to trust. I asked for meaningful financial compensation, a written incident report, and proof of the backup schedule and retention in place.

At 8:15 PM on 7 October the support manager offered one month of free hosting and called it the final offer. The same day, a generic message said the compensation claim had been escalated for further review.

At 8:11 AM on 8 October I rejected that too. I asked for at least 24 months of free service, a written explanation of why xyz_vynex was not in the backups, and an explanation of why my server had been suspended without an invoice being sent.

8 October: the suspension and the $144 invoice

Phase 7

This is the part of the story I still cannot fully explain from the ticket.

I say the server was suspended for non-payment, and that I never received an invoice before the suspension. In a second ticket, #2026100886572, opened on 8 October, an agent noted the suspension reason as non-payment (an overdue invoice) and wrote that "the customer states that they cannot see any unpaid invoice in their Client Area." The same ticket shows a renewal invoice, #917078, generated at 8:04 AM on 8 October. The thread does not show when the suspension started.

The licence invoice I paid at 6:05 PM on 5 October is a separate invoice from this renewal and is not part of the $144 discussed here.

The invoice was for $144: $119 for the server plan and $25 for a cPanel Admin licence for five accounts. I had asked for Solo at $15 and I had said in writing that I needed one account. The $25 Admin line was the billing error. The billing agent confirmed the breakdown at 8:33 AM. At 8:56 AM I wrote that I had paid under protest to get my business back online, and I asked for three things: the licence changed to Solo, a $10 refund to my original payment method, and written confirmation that future renewals would be $134.

At 9:24 AM billing said it could not downgrade the licence or refund the current cycle, "as the renewal has already been processed and the license is now active", and that Solo could be applied at the next renewal. At 9:58 AM I replied that the mid-cycle rule did not apply, because I had not changed my mind; the Admin licence was a billing error, and I had asked for Solo six days earlier. I said I would dispute the $10 with my payment provider if it was not refunded.

The ticket page for the first case shows its state at 10:01 AM on 8 October. UltaHost wrote that the 5-account Admin licence had been billed in place of the Solo licence I requested, and that the team was reviewing "the licence change and the 10 USD overcharge, along with the suspension concern, as a priority."

What it cost me

I am not going to pretend the numbers are settled. Some are in the ticket. Some are my own estimates. I have marked which is which.

• Downtime. The ticket shows the site in maintenance mode from the afternoon of 5 October. The restore was reported complete at 11:41 PM, but the production database was not in it. I reported "over 21 hours" of downtime at 8:09 AM on 6 October and "about 20 hours" in later messages.
• Data. The database xyz_vynex held every customer account and order for vynex.ai. The period between the 30 September backup and the 5 October reinstall is the window where new customers and orders were lost. UltaHost confirmed the database was in the 30 September backup. I then recovered it myself.
• Customers. I reported more than 300 registrations in the preceding four to five days that I then had to explain to customers.
• Search traffic. My own figures, from my analytics and Search Console for the days after the outage: daily traffic fell from around 50,000 visits to between 15,000 and 18,000, and key Google rankings dropped from positions 1 to 3 to positions 7 to 10. These are not shown in the ticket screenshots.
• Revenue. I estimated more than $400 in lost ad revenue in the first 20 hours alone, and total losses above $20,000 as of 8 October.
• Extra cost. I paid $144 for a licence I had not ordered, under protest. I am asking for the $10 difference back as a refund to my payment method.
• Time. I spent roughly a day after the restore rebuilding the server: cron jobs, the PHP version, mail, and the admin tools.

What the ticket proves and what is still my claim

A reader should be able to separate the documented facts from my own account. Here is the split.

Claim	What the ticket shows
The server was reinstalled without a full-disk backup first.	Confirmed. UltaHost's 2:36 AM message on 6 October says the latest full-disk backup was dated 30 September. The restore used the cPanel account backup only.
The database xyz_vynex was missing from the restored cPanel account.	Confirmed by UltaHost on 6 October at 2:36 AM and 6:47 AM.
UltaHost should have verified the backup before reinstalling.	Admitted by UltaHost on 6 October at 6:47 AM: "we should have verified that it was backed up before reinstalling."
The database existed in the 30 September backup.	Confirmed by UltaHost on 6 October at 7:36 AM ("We found xyz_vynex in the September 30 backup").
The server was suspended for non-payment without an invoice being sent.	Partly confirmed. The 8 October agent note records the suspension as non-payment and says the customer cannot see an unpaid invoice. A renewal invoice (#917078) was generated on 8 October at 8:04 AM. The thread does not record when the suspension began, so the start date rests on my own account.
I was billed for a 5-account Admin licence after asking for a Solo licence.	Confirmed. The $144 invoice shows a $25 cPanel Admin (5 accounts) line. My 9:53 AM message on 2 October says a Solo account is enough.
Downtime of about 20 hours and traffic and ranking losses.	My own figures. The thread shows the maintenance start and the restore completion time, not total downtime. Supporting data is listed above in the impact section.

What every hosting buyer should check first

I did not expect to write a checklist, but this incident is the reason I now run one before any server work. Every item below would have prevented or shortened this outage.

1. Ask for a full-disk backup before any reinstall, and check it yourself. A cPanel account backup is not a server backup. Ask the host to confirm which databases are in the archive, and check one table of your own database in it before the job starts.
2. Keep your own offsite database dumps. A nightly mysqldump copied to storage you control is the only backup you fully own. Test a restore once a month.
3. Get the procedure, the timeline and the rollback plan in writing. Ask what happens if the restore fails halfway, and what the host will do about it.
4. Confirm the licence tier against what you ordered before paying. Ask for the invoice line items and check the account count.
5. Ask whether a renewal invoice is pending before work starts. A suspension should never be the first sign of a billing problem.
6. Get a time-stamped ETA and hold the host to it. When an ETA slips, ask for the new one in writing every hour rather than waiting for the next update.
7. Keep your own uptime monitor running. If your monitor shows the outage start time, you do not have to rely on the host's account of when it began.

Where things stand

As of the last entry I have on the first ticket (8 October at 10:01 AM), UltaHost had acknowledged the wrong licence, the $10 overcharge and the suspension concern, and said they were reviewing them as a priority. There is no agreed compensation figure. My rejected offers were two weeks, then one month, of free hosting. I have asked for:

• the $10 difference refunded to my original payment method, and the licence changed to cPanel Solo;
• written confirmation that future renewals will be billed at $134 per month;
• a written incident report explaining why the production database was not in the backup, and what has changed so that it cannot happen again;
• written proof of the backup schedule and retention on my server;
• compensation that reflects the downtime, the lost data and customers, and the lost traffic. I am open to discussing a figure in writing with someone who has authority to approve one.

I would rather settle this directly than in a chargeback or a public dispute. This page is the record that I keep. It reflects the ticket as of 8 October 2026 at 10:01 AM and will be updated if anything changes.

Written by Moadood Ahmad, founder of vynex.ai. Sources: the support ticket pages and screenshots captured by the author, 2 to 8 October 2026. Figures marked as the author's own estimates are not independently verified.

Frequently asked questions

What exactly went wrong with the UltaHost cPanel reinstall?

UltaHost reinstalled the server using a cPanel account backup. The production database was not mapped to that cPanel account, so the backup skipped it. When the restore was reported complete on 5 October 2026, the site had no database. The newest full-disk backup UltaHost held was from 30 September, so six days of customer accounts and orders were missing.

Did UltaHost admit fault?

Yes, in the ticket. On 6 October 2026 at 6:47 AM a support agent wrote that the cPanel account backup did not include the database and that they should have verified it was backed up before reinstalling. UltaHost later confirmed the database was present in the 30 September full-disk backup and provided a temporary server to extract it from.

How long was the site down?

The site went into maintenance mode on the afternoon of 5 October 2026. The restore was reported complete at 11:41 PM, but without the production database. The temporary recovery server was made available at 8:30 AM on 6 October, after which the author exported the database himself. The author's own figure for total downtime is about 20 hours, followed by roughly a day of server repairs.

What compensation did UltaHost offer?

UltaHost first offered two weeks of free hosting, then one month, which it called its final offer. Both were rejected. As of 8 October 2026 UltaHost had acknowledged the wrong licence tier, the $10 overcharge and the suspension concern and said it was reviewing them. No compensation figure had been agreed.

What should I check before letting a host reinstall my server?

Ask for a full-disk backup, not just a cPanel account backup, and confirm your databases are in it before the work starts. Keep your own offsite database dumps and test a restore. Get the procedure, timeline and rollback plan in writing, confirm the licence tier on the invoice, ask whether any renewal is pending, and run your own uptime monitor so you have independent outage timestamps.