All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
HostBRR support has been silent for months - Storage VM keeps getting slower
I've been a HostBRR customer since May 2025, on the EPYC Storage - 1TB - Germany plan at €3.50/month. All 17 invoices have been paid on time.
The VM itself has been fine. No uptime issues that i can remember and performance is what I expected for the price. I only use it as a small relay and secondary backup box tho, so I don't push it hard at all (cpu usage avg %1.1 in last 120 day.)
The problem is the storage disk volume. It keeps giving problems to me. I run the same fio test (4k random, QD1) in three different times:
| May 2025 | 12 Oct 2026 | 13 Oct 2026 | |
|---|---|---|---|
| Read IOPS | ~1,460 | ~130 | ~72 |
| Write IOPS | ~1,460 | ~130 | ~74 |
| Read speed (4k random) | ~5.9 MB/s | ~0.53 MB/s | ~0.30 MB/s |
| Write speed (4k random) | ~5.8 MB/s | ~0.54 MB/s | ~0.30 MB/s |
| Median read latency | ~0.05 ms | ~0.14 ms | ~7.8 ms |
| Sequential write (1M) | ~24 MB/s | ~14 MB/s | ~21 MB/s |
| Sequential read (1M) | — | ~319 MB/s | — |
Writes also stall for up to about a minute now and then, and sequential writes swing a lot from day to day. The system disk on the same VM is fine, so it's specific to the data volume.
I sent the May 2025 fio results to support at the time, and it got a bit better. But it repeats again.
The main issue is that their support. I have a ticket #9358385 ("Affiliate payout") that had no reply for about 144 days. I also opened another ticket for disk speed 38h ago #4702123 ("about hybrid disk speed") and also has no reply yet.
I also sent a forum DM to @labze on 4 August for the first ticket above and got nothing back/ So I'm posting here as a last resort.
I know it's a budget server, and I don't expect much, but it's just getting more and more annoying at this point.
Comments
share your fio test script so i can check too.
Their disk speeds are horrible. I'm definitely not renewing.
And it's not like I ask much, but 12MB/s write is just not worth it for me to keep.
That is very bad. I do apologize!
I'll look into this shortly.
Response times are a bit hit and miss. There has been a overwhelming demand lately causing a backlog of tickets.
Well it is a mistake to test random 4K operations on a shared HDD array. We intentionally limit IOPS to prevent a few clients from degrading the entire array.
I mean, even a dedicated HDD chokes on random 4K IOPS tests.
OP, what is your backup method? If using large, compressed, files, this isn't an issue. If you're rsyncing 10k individual files, you're doing it wrong.