New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Do you have any special promotions planned? It looks like you’re almost at your 5th anniversary.
Nothing planned yet. I'm sure we'll do something, but what it is, haven't decided.
Will cross that bridge when I get there.
PS:
Support has been better recently.
Last week's stats:
UPDATE (for reference of others):
Before booting the VM on the VirtFusion hypervisor after migrating the disk image over, first run:
IBGUESTFS_BACKEND=direct virt-sparsify --in-place disk-file.imgI happened to notice earlier that every migrated VM was reporting as 95%+ disk usage in VirtFusion whereas the original disk usage shown in Virtualizor was often much much less. On our internal VMs I migrated over I logged in and saw that this was incorrect by actually checking the real usage via SSH.
A couple minutes on @VirtFusion 's Discord and Phil responded with what to do to fix it. Great support, and yet another reason why VirtFusion is great. (Virtualizor would have asked for credentials first, almost certainly)
Couple thoughts on this.
In the qemu-img conversion stage, you should be able to sparsify in that step (man qemu-img, I think it's -S and you set a block size.)
You could also skip the rsync step in a couple of ways:
cat or pv the image, pipe into SSH, and have the VirtFusion host do the qemu-img conversion step. (someone clever might add a gzip -1 and gzip -d to this pipeline to speed things up over zero blocks)
qemu-img convert into SSH, and then do the sparsify on the host.
Now I don't think you can write sparse over an SSH pipe, but rsync should support copying sparse. But I think for rsync it's a manual flag.
I would personally do the qemu-img conversion on the destination host side and convert while you stream it. Should be pretty straightforward. You'll definitely want to gzip it as it will make most disks go much faster.
The other side of the coin:
If you care about sparse allocaiton, are you overprovisioning disk? I am not a fan of disk images on the filesystem in general, though it certainly does work. OpenBSD's VMM only supports reading from a qcow or raw file on the filesystem. For me in this case, I explicitly write out every zero and do nothing sparse. This is partly because most of my hosts are on spinning disks, but either way avoiding fragmentation can be a good idea.
Now if you do fully allocated qcow with nothing sparse, the only benefit is if you want to do snapshotting. Which can definitely be worthwhile. And in some cases, the sparse qcows are easier to backup. In my case, VMM doesn't support live snapshotting of any kind, so it's a moot point and I just do raw disk images.
(You'll see the best performance out of something like LVM or ZFS volumes, but you have to be careful to zero out your space after use. Maybe not with ZFS, but definitely LVM. Both will give you snapshotting capabilities. I am not sure how trustworthy ZFS on Linux is. Maybe it's pretty good these days? This is one thing that has made me curious about FreeBSD as a hypervisor, specificially the trustworthy ZFS aspect.)
Try
zstd -c --adaptinstead. Not only is Zstandard a much faster algorithm than DEFLATE, but zstd's adapt mode will dynamically adjust compression strength based on available I/O to maximize compression ratio without becoming a bottleneck. And zstd, unlike gzip, is multi-threaded.It's rock solid. All issues it has are simple QoL issues like ARC poorly integrating with the rest of the memory management subsystem (so it makes it look like the reclaimable cache is not reclaimable, even though it will auto-shrink), unnecessary duplication of the page cache for
mmap()ed files, etc. In sheer terms of reliability, it's the most trustworthy filesystem you can run. Although I hear bcachefs is fighting for that throne, but it's not ready for prime-time yet.Can confirm, Linux ZFS is more than mature enough now. I have been using it for over five years now including for VM .raw disks. It wasn't even any trouble to take my BSD ZFS pools onto Linux when I got bored of BSD itself giving me trouble.
Btrfs is really the one to still be a bit skeptical about, but even then I run it on one of my machines and it hasn't been a problem yet, I just know it has some "history" of not being that great.
Happy 5th Birthday incognet
Happy birthday!
Happy birthday! 🇹🇭
Happy Birthday!
happy birthday @manndude. thank you for caring about privacy and doing what you do
Thanks but y'all are a month early
I opened yet another ticket regarding SEV-ES availability in Bulgaria. Previous one was closed without any response around 4 months ago. I would greatly appreciate if you could review the possibility of enabling this feature
Currently, I have no way to verify that IncogNET actually supports SEV-ES anywhere.
were you ever able to get SEV enabled?
Not yet.
As it currently stands, IncogNET has replied to my ticket shortly after I posted here.
They informed me SEV may not be tested/enabled in Sofia, Bulgaria (where my VPS is) yet and proposed to move me to Liberty Lake, Washington (USA) where SEV-ES is supposedly supported.
Of course I will give updates here how this works out
Ok thanks for the update, I wish the locations would be clearly marked with their features on the homepage vs having to dig around for the info! Let us know how it works if you decide to transfer location
Sorry for the delays. Got news about a parent diagnosed with a terminal illness on Monday of this week so have been trying to make a plan for what I need to do in regards to that. Watching your parents get older brings more heartbreak than joy.
If we get the SEV stuff available as a quick-toggle on/off option internally from our side we may display it as an advertised feature again, but since not all hardware deployments support it it now may just be a "If you know about it, and ask about it, we may do it. We may not, especially for a cheap promo server." type thing. Not really wanting to have tickets about it, having to do migrations and test things boot up properly for a $12/yr 512MB VPS or something. It's super niche and even when it was temporarily displayed as an option, had nearly no interest from anyone. I guess I was more optimistic at one point about it. Sort of like our old Tor and I2P clearnet mirroring for shared hosting. Advertised it for a couple years. Probably had half a dozen shared hosting customers opt in for it. Was used so little that it often became a point of pain during upgrades and updates or when needing support. (Though that is likely to come back, spent a couple weeks last year automating things a bit for this and testing out a newer, better way to complete this than how it was done before.)
Anyhow, the amount of tasks I have on my to-do list at this moment exceeds my energy and focus levels as well as the hours I have reasonably in a day to work. Just the normal day-to-day stuff takes hours and I'm no longer in a position where I can sit at my desk for 12-16 hours a day 6 or 7 days a week like I used to.
Right now I'm still doing Washington legacy migrations to VirtFusion. I think I have about 2 hypervisors left to migrate but I only do about 10 VMs a night since it's a manual process and slow going having to check things. PA, KCMO and NL will come sometime after. Assuming no more hardware failures in WA like I had the other day (when it rains it pours) I'll get caught up on what I consider "extra curricular tasks", especially those that require me to dig through my old notes and stuff. Not forgotten about, I swear!
To curb some abuse, this week I've made the 1GB RAM plans available with quarterly options available as the minimum billing period. The 512MB plans remain available with a semi-annual option being the shortest billing period. This may be temporary, we'll see.
Soon, new customers only, who pay with PayPal will have stricter requirements. Especially for larger orders. PayPal email address will need to match the order email address, 100%, as a minimum. Email verification for PayPal orders may be required. Tired of trash wasting time and energy trying to do chargebacks, of which we've won every case this year, by the way, but not at the expense of having to document account details, order details, and correspondence with PayPal. Seems to be mostly issues with our Bulgarian POP, which is the same POP that has forced us to reduce the minimum billing period for the cheaper plans. This won't impact existing users, only new ones. Use crypto, mail forwarding/relay and/or a VPN or Tor if you want max privacy. (Or just use PayPal and be an honest, halfway decent human being)
OMG, I guess there is nothing much anybody can say in the light of your personal news. Well, I just wish you and your parent mental strength.
Damn, that's some sad news
I'm really sorry you’re going through this. Remember we are always here for you, in case you need to collect thoughts or just talk.
That is not an issue. It's fine to not support it as well and just admit it's not available. Totally understand the part about it being niche and not easy to have supported.
First of all, I hope everything turns out for the better regarding your family situation. I completely understand that that takes priority.
Just to clarify, does this mean the account email and PayPal email need to be the same? What is “the order email address.”
And which email will be verified for PayPal orders?
I ask because I use a different alias between PayPal and Incognet.
(Also tbf I’m not a “new customer,” but it’s probably helpful to prospective customers to ask).
Basically, order email must match paypal email, if it doesn’t it will most likely send you a code to verify you own it.
I have did that before, but too much hassle, did not have a chargeback for few years already.
Are existing users defined as those who already signed up, or those who signed up and submitted their first payment? Some are up and running but haven't had to pay yet because of a promotion.
I am a fairly loyal customer of Incognet of 4 yrs (and I kept my VPS through the shitshow that was Amsterdam migration) because I genuinely like what Incognet is trying to do and when service works - it's amazing. Conversely, when it doesn't - support seems to be barely or non-existing.
Context:
Paid my invoice on Monday 2026-04-06 12:00 UTC [admittedly, I paid it after third and last overdue notice, which is my screw-up]. After checking few hours later I notice that my invoice is still in "Unpaid" state. Not a big deal, I think to myself and open a ticket #0406K68A5.
Two days pass with no reaction whatsoever on my ticket, and I send a message to @MannDude on Wednesday with a request to take a look at a ticket.
Still no reaction.
Somewhen in-between my pm to MannDude and now my VPS goes into Suspended state.
Now, it's Sunday evening. My VPS is still in a suspended state, despite me paying invoice almost a week ago.
I assume you paid via alt-coin?
BTC, LTC, DOGE and XMR are processed in house and often without issue. Anything else is via Trocador, but lately some invoices are expiring before payment completes or being "partially settled".
In anycase, checking now. Billing stuff is my responsibly as the only one with that level of access so I do apologize for the delay.
EDIT: Should be good now. Tron -> Trocador -> IncogNET caused the invoice for to expire (120 minutes) so the received amount only "Partially Settled" the invoice. The "still due" amount that was preventing BTCPayServer from settling the invoice and kicking off the webhooks to WHMCS was just
0.00000101 LTC... (Alt-coins we accept are converted to LTC for our use.)I've adjusted the settings so
Consider the invoice paid even if the paid amount is … % less than expectedis more appropriate for minor fluctuations in the received amount versus the expected amount.You've been given two months of service for the inconvenience.
How's Sweden these days, did the rough edges end up smoothed out?
Yep, Sweden is working fine for me now.
All good in Stockholm.
Sweden has been solid again.
Maintenance announced in WA for Monday, emails out now. Sorry for short notice, I just got confirmation of the timeslot about an hour ago now.
Working on a big DNS update as well. MyPrivateInbox, so close. But not launching anything new until some more important things are taken care of first and some better monitoring and workflow optimizations get implemented for existing stuff so that we can be one step ahead of and potential problems.
Sort of got behind again on my own tickets and tasks flagged for me with some personal life stuff + New Years stuff (SE Asia does it in April, for three days 🤷) but still bashing it out. If something is waiting, it's likely due to me personally. Going to get back to it this weekend and caught back up.
I've updated my coffee game from a French Press to a Moka Pot. It's what's been fueling me. Took my manual coffee grinder apart for cleaning and discovered that the mechanism for grinding can be adjusted for coarseness, so now I'm getting my beans ground exactly how I want them. I swear I'm not a coffee snob though... 👀
Mmmm!! Yummy, yummy, yummy! Can’t get enough