There's no way I'm touching that. I'm not very good with trying to get it to work in the first place so it's a miracle. I think I spent probably an hour on that and it was bouncing all over the place. This is as close to normal as I can get it unless someone wants to volunteer an improvement.
I'm guessing that's on a specific resolution I didn't test. Oh yeah it's if the screen is smaller, replicated. Get a bigger monitor and fullscreen it so I don't have to fix it. Or make it even smaller.
Oh it's based on height size not width. Let me see... okay fixed by disabling it based on height as well. That was easy.
Some people talked about notifications earlier and I had something like this written out but I'm trying to compile a list of improvements wee should make while keeping them realistic.
Try to make backend notifications less excessive, you likely still won't be customize them (very much, if at all) but I might be able to add some "agreed upon" constraints.
I'm still not sure if it's just mine that's bugged out (since I was testing it and didn't reload it to clear my settings) and haven't had time to check any further, but is anyone else receiving the BACKEND notifications while the frontend is open? If so I need it reported as that's not intended and it would need a fix.
Fixed a bug where the notification SOUNDS (not TTS) wouldn't fire immediately if you had it tabbed out and all fired at once giving you a scare once you tabbed back in. This was older logic for a way we were going to do it that I didn't update.
I might try to add SOME level of control for backend versus frontend notification let you disable one or the other instead of both being on or both being off.
After these changes, if you guys have backend notifications enabled, you'll have to visit the page again to refresh your ... whatever it's called, subscription? I'll give an announcement once that's done (and I'm done messing with it.)
@localhost said: @VirMach in the notifications can you add the deal number? Seems I am getting either phantom notifications or too late.
I'll also try to also add in the plan number as requested for the backend notifications. I need more information on it being late or "phantom" because it matters if it's the backend one, frontend, etc. Also some of it out of our control. It depends on your browser, your operating system and how they both manage notifications. They've made a lot of "smart" changes where they hide them or delay them based on your patterns and I don't believe we can override that. I might also add a timestamp, probably in UTC, for the backend ones as those seem to be delayed like that more than the frontend.
One of the "smart" features I think we could take advantage of "taking back" thee previous notification and/or updating it. I believe the way we have it set up is a unique notification per plan ID, but we could possibly clean it up to where it goes based off each product card, if I can figure that out. Then that way it should get updated and you won't end up with like 20 notifications on your notification sidebar.
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
@tenpera said: @VirMach, now it's #165, when will it go back to #1?
It's not an odometer you can roll back, it just goes up to infinity.
What happens when the dealz number reaches Number.MAX_SAFE_INTEGER?
JavaScript will lose precision, Tokyo will become Bangkok, FrankZ & Wonder_Woman will be sucked into eno1 & eno2 and get married.
What happens after the heat death of the universe, will my discount still be recurring?
@VirMach said:
Some people talked about notifications earlier and I had something like this written out but I'm trying to compile a list of improvements wee should make while keeping them realistic.
Try to make backend notifications less excessive, you likely still won't be customize them (very much, if at all) but I might be able to add some "agreed upon" constraints.
I'm still not sure if it's just mine that's bugged out (since I was testing it and didn't reload it to clear my settings) and haven't had time to check any further, but is anyone else receiving the BACKEND notifications while the frontend is open? If so I need it reported as that's not intended and it would need a fix.
Fixed a bug where the notification SOUNDS (not TTS) wouldn't fire immediately if you had it tabbed out and all fired at once giving you a scare once you tabbed back in. This was older logic for a way we were going to do it that I didn't update.
I might try to add SOME level of control for backend versus frontend notification let you disable one or the other instead of both being on or both being off.
After these changes, if you guys have backend notifications enabled, you'll have to visit the page again to refresh your ... whatever it's called, subscription? I'll give an announcement once that's done (and I'm done messing with it.)
@localhost said: @VirMach in the notifications can you add the deal number? Seems I am getting either phantom notifications or too late.
I'll also try to also add in the plan number as requested for the backend notifications. I need more information on it being late or "phantom" because it matters if it's the backend one, frontend, etc. Also some of it out of our control. It depends on your browser, your operating system and how they both manage notifications. They've made a lot of "smart" changes where they hide them or delay them based on your patterns and I don't believe we can override that. I might also add a timestamp, probably in UTC, for the backend ones as those seem to be delayed like that more than the frontend.
One of the "smart" features I think we could take advantage of "taking back" thee previous notification and/or updating it. I believe the way we have it set up is a unique notification per plan ID, but we could possibly clean it up to where it goes based off each product card, if I can figure that out. Then that way it should get updated and you won't end up with like 20 notifications on your notification sidebar.
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
@tenpera said: @VirMach, now it's #165, when will it go back to #1?
It's not an odometer you can roll back, it just goes up to infinity.
What happens when the dealz number reaches Number.MAX_SAFE_INTEGER?
JavaScript will lose precision, Tokyo will become Bangkok, FrankZ & Wonder_Woman will be sucked into eno1 & eno2 and get married.
What happens after the heat death of the universe, will my discount still be recurring?
Order #611314 It's almost 24 hours. It still hasn't turned on.
Any other feedback or suggestions or requests you have on how things should work, speak now so we have time to include it in potential changes.
@mucstudio said: Order #611314 It's almost 24 hours. It still hasn't turned on.
This is because of an error catcher, if it catches an error in configuration settings it rolls it back and doesn't activate it. Before it would activate it with the wrong specifications and I'd manually fix it. Once the code's fixed it should generate. I might remove the validator soon so they can just generate (wrong) and I can go through and fix them versus just being stuck. I'll decide in a few hours.
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
EDIT: Can be installed from solusvm but not from customer panel (Virtual server not found)
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
Yeah same thing, this is actually not a vserverid error so it's fine. It's just marked as active but not actually active. I might add for it to go back to pending status so people don't think it's broken but it's better to just fix it so it validates.
@remy said: Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
EDIT: Can be installed from solusvm but not from customer panel
I fixed the logic, but the rollback logic had an error. So yours actually just created but rolled back. I don't know what happens now, definitely don't use it on SolusVM right now, it might get terminated.
Either it will re-create now and old service will stay there and manually cleaned up/terminated, or it'll get permanently stuck until I manually go through these that created a few minutes ago and fix them.
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
Yeah same thing, this is actually not a vserverid error so it's fine. It's just marked as active but not actually active. I might add for it to go back to pending status so people don't think it's broken but it's better to just fix it so it validates.
@remy said: Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
EDIT: Can be installed from solusvm but not from customer panel
I fixed the logic, but the rollback logic had an error. So yours actually just created but rolled back. I don't know what happens now, definitely don't use it on SolusVM right now, it might get terminated.
Either it will re-create now and old service will stay there and manually cleaned up/terminated, or it'll get permanently stuck until I manually go through these that created a few minutes ago and fix them.
Ok no problem. I'll just wait your fixes
I was deployed on the same node as my previous VM, no luck
@remy said:
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
Yeah same thing, this is actually not a vserverid error so it's fine. It's just marked as active but not actually active. I might add for it to go back to pending status so people don't think it's broken but it's better to just fix it so it validates.
@remy said: Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
EDIT: Can be installed from solusvm but not from customer panel
I fixed the logic, but the rollback logic had an error. So yours actually just created but rolled back. I don't know what happens now, definitely don't use it on SolusVM right now, it might get terminated.
Either it will re-create now and old service will stay there and manually cleaned up/terminated, or it'll get permanently stuck until I manually go through these that created a few minutes ago and fix them.
Ok no problem. I'll just wait your fixes
I was deployed on the same node as my previous VM, no luck
Should have finally provisioned correctly. Well almost, I'm updating it manually to the final product. Looks good now.
Just another reminder on duplicate services - don't use the one that's not on WHMCS. I mean you could use it, I don't care, but the duplicates during the BETA will likely have NO grace period and NO support in switching back to it, recovering data, etc. If we forget to clean them up, I don't want a ticket several months from now asking to grab data off the one that wasn't on WHMCS.
Oh wait I forgot I have a tool for this, I'm going to run it now for the newly created ones so we can eliminate that confusing possibility.
@remy said: Ok no problem. I'll just wait your fixes
I was deployed on the same node as my previous VM, no luck
I've ran it now, the one(s) you shouldn't use are suspended and the correct one is active. Let me know if anything else looks weird since yours got created through like 3 or 4 code changes. Basically report anything weird immediately and definitely keep backups for any super important files around 2 weeks per the recommendation (or honestly, you should always do that anyway.)
@VirMach said: Any other feedback or suggestions or requests you have
It'll be nice to have 3 different locations to pick from (along with the different specs/prices of course).
Also, for the crazy people here (myself included), since you've risen from the ashes (yet again?), can you:
a) Reward those of us who have renewals (and still have faith in you for whatever weird reason, despite not even knowing where the instance is running or if it is in existence at all - does a VPS that doesn't answer pings exist at all)?
b) Some recurring coupons please? So far everything that I've seen is a one off (very generous at that of course) but it'll be nice to encourage some longer term customers who would like to stick with you (or should I say tolerate) for a year or two more, especially if the price is attractive?
c) Can you throw the old faithful a few hidden gems and treasures for a seek-and-you-shall-find kind of a game? You know we're suckers for hard to get things at low prices even if they are not going to be used beyond the trophy-I-snagged-it stage.
d) How about a nice "bundle" (or a pick any 3-5 locations type of thingy) - same (decent low end specs 512MB or 1GB of RAM, 1 or 2 vCPUs, 2-5 TB of BW, 25-40GB disk, 1 IPv4 + IPv6... yeah I know you don't do IPv6 but come on... it's 2024 and you've been promising for a few years now and finally you're out of clutches of that DC that doesn't do IPv6), all locations
OK. I've run out of ideas other than creating a cheeky VeerBot that veers off the specs every now and then and shocks you with some insane offer.
If you've read this far, surely it was entertaining enough and you should throw us an offer that we can't refuse.
@nullnothere said:
c) Can you throw the old faithful a few hidden gems and treasures for a seek-and-you-shall-find kind of a game? You know we're suckers for hard to get things at low prices even if they are not going to be used beyond the trophy-I-snagged-it stage.
Let's just say it might or might not have already happened...
@nullnothere said: It'll be nice to have 3 different locations to pick from (along with the different specs/prices of course).
I think if we did any work to get this done, it might actually end up being worse than how it's done now. The current set of offers that are about to go live (they just recently expired as I'm writing this) are Tokyo, Amsterdam, and San Jose. If you look at every chunk of 3, there's usually at least two different locations. If we force it to keep plans in different locations, some locations would end up running out sooner and making the future diversity worse, and the plan diversity would also suffer. I feel like randomized is the most interesting pattern.
@nullnothere said: a) Reward those of us who have renewals (and still have faith in you for whatever weird reason, despite not even knowing where the instance is running or if it is in existence at all - does a VPS that doesn't answer pings exist at all)?
Those who stick around are naturally already skewed toward cheaper services, so it's difficult to provide any type of discount on top of that. Also your statement, while true in some very specific clusters/cases, is not anywhere close to the actual big picture. You're focusing on let's say 5% of all cases. The other 95% is mostly several hundred days of uptime. We've also already been (usually) generous with those major outages/issues with service extensions. If you point out specific cases, I can take a further look and make sure that people were appropriately credited.
@nullnothere said: b) Some recurring coupons please? So far everything that I've seen is a one off (very generous at that of course) but it'll be nice to encourage some longer term customers who would like to stick with you (or should I say tolerate) for a year or two more, especially if the price is attractive?
I know there's a psychological thrill to having prices be higher and then receiving a discount on top of it, but take a look at the specials as they are currently. They already get further price cuts automatically in most cases if they hang around too long, and on the lower end you can easily tell that the price formulas are set up in a way where they barely even cover IPv4 costs. There's a calculated percentage off on all of them, and that's pretty accurate compared to our regular plans, with most of them being in the 80% or higher discount range, some hitting like 88% or more. So with that said, I'm unsure how much more you want off recurring to consider it to be "attractive."
Perhaps I haven't looked around much recently. I know you're setting up a lot of this as "VirMach's intolerable" so my next question might have the answer that "VirMach is intolerable so it has to be cheaper"... but are there any other providers with active offers, on newer hardware, that are providing deals with an IPv4 for lower than $6 to $8 a year?
By the way the idea behind the one time discount isn't to try to encourage more sales, it's just a way to give people the first few months free because something might go wrong and the service might get stuck/unusable for a while. That hasn't really happened but it was just a precaution.
@nullnothere said: c) Can you throw the old faithful a few hidden gems and treasures for a seek-and-you-shall-find kind of a game? You know we're suckers for hard to get things at low prices even if they are not going to be used beyond the trophy-I-snagged-it stage.
We're barely back to being able to do a sale, and even now, there's other outstanding issues and not a lot of time. I wish I could, it's fun for me as well, but the second issue is that as I mentioned earlier, renewals are skewed toward these harder finds and over the last 5-10 years, the numbers of them have gone up, and they're now a higher percentage of all total services. So we have to be careful about the quantity. I'm not saying there won't be any, but the quantity that would be available would be so rare that it's likely not worth the time & effort for anyone this time around.
@nullnothere said: d) How about a nice "bundle" (or a pick any 3-5 locations type of thingy) - same (decent low end specs 512MB or 1GB of RAM, 1 or 2 vCPUs, 2-5 TB of BW, 25-40GB disk, 1 IPv4
This might be possible but it wouldn't be a bundle with flash sales. It'd have to be some other set bundle.
IPv6... yeah I know you don't do IPv6 but come on... it's 2024 and you've been promising for a few years now and finally you're out of clutches of that DC that doesn't do IPv6), all locations
I really do want to get IPv6 going but SolusVM support for it is abysmal. I don't feel comfortable doing it and if we did it I would want to not introduce more problems. We just haven't had the time to finish the system required to do this with confidence. I don't want to give any timelines because as you said, it's been a while. It'll happen eventually.
And yes it's kind of ironic how much we even personally made fun of that DC that doesn't do IPv6 and we kind of ended up turning into them.
@VirMach said: Any other feedback or suggestions or requests you have on how things should work, speak now so we have time to include it in potential changes.
I am blind. Didn't read your wallll of text for deal numbers in notifications.
I think if we did any work to get this done, it might actually end up being worse than how it's done now. ...
If we force it to keep plans in different locations, some locations would end up running out sooner and making the future diversity worse, and the plan diversity would also suffer. I feel like randomized is the most interesting pattern.
Oh. I didn't think this through from your perspective (esp. varying stock levels vs popularity). Fair enough.
The other 95% is mostly several hundred days of uptime.
I'd like to think so, but I have also ended up pruning my "collection" to a bare minimum because I lost track of where what was and if it was running at all (through all the migrations and what not). So I'm not very clear on what happened (or why). I'm now going to look at a few low end offers via the VirBot and maybe pick a few here and there just for kicks. I've mostly abandoned hopes of a big powerful VM somewhere because of the huge migrate-away-from-the-evil-DC fiasco whenever it happened (see - I've lost track of time as well.
I'm unsure how much more you want off recurring to consider it to be "attractive."
It's mostly looking at balanced offers from the VirBot vs other BF offers around. Something like a 2-4 vCPU/4GB RAM/60GB Disk/5TB BW at roughly 20-25 USD (or EUR, stretch a bit) equivalent is very reasonable and attractive and I'd even hazard reasonably "frequent" during BF across locations. Add a no-ipv6 penalty/discount (for those of us who do care about IPv6) and
I think you're entering pretty decent sale territory.
...but are there any other providers with active offers, on newer hardware, that are providing deals with an IPv4 for lower than $6 to $8 a year?
Surprisingly there are a couple who are in that price range (and already have sold out) - but let's be realistic, no one is looking for a mass deadpool offer. A few sparse crazy offers to up the ante and the regular balanced sale offers (hopefully sustainable) will be what people are looking for mostly.
By the way the idea behind the one time discount isn't to try to encourage more sales, it's just a way to give people the first few months free because something might go wrong and the service might get stuck/unusable for a while.
That's fair (and appropriate) - I didn't think that way at all (see, I'm thinking only "consumer" side not "producer" side)
We're barely back to being able to do a sale, and even now, there's other outstanding issues and not a lot of time. I wish I could, it's fun for me as well, but the second issue is that as I mentioned earlier, renewals are skewed toward these harder finds and over the last 5-10 years, the numbers of them have gone up, and they're now a higher percentage of all total services. So we have to be careful about the quantity. I'm not saying there won't be any, but the quantity that would be available would be so rare that it's likely not worth the time & effort for anyone this time around.
Agreed and again, not a perspective I had considered. I do understand on the outstanding issues which we do want fixed (and that is why I think this thread is popular - you're responsive or at least good at faking it!) - and I'm also quite happily engaged because you did take the time to put together a rather long, well thought out, point by point justification of your stance (note I didn't use the word rebuttal here!).
I really do want to get IPv6 going but SolusVM support for it is abysmal. I don't feel comfortable doing it and if we did it I would want to not introduce more problems. We just haven't had the time to finish the system required to do this with confidence. I don't want to give any timelines because as you said, it's been a while. It'll happen eventually.
Well... how come so many of the other providers (many of whom can't even get emails to work properly) are able to provide good IPv6 functionality with or without SolusVM? You're literally the ONLY provider I have VMs with that don't have a proper functional IPv6 setup (at most locations) but but but because it's VirMach, I'm kind of still OK with a few IPv4 only test vms (again, because it's cheap enough, mostly at the <$10 price point, that I don't mind "splurging"). But it's getting to be an irritant and I really hope you can at least make some sort of a fix for the crazy knowledgeable ones to use IPv6.
Thank you for taking the time to go though and put together your points.
Now where is that reward for such a flawless work of literature on the merits of engaging nicely with a provider to ask for a shameless discount coupon that lasts to eternity and beyond?
@nullnothere said: It's mostly looking at balanced offers from the VirBot vs other BF offers around. Something like a 2-4 vCPU/4GB RAM/60GB Disk/5TB BW at roughly 20-25 USD (or EUR, stretch a bit) equivalent is very reasonable and attractive and I'd even hazard reasonably "frequent" during BF across locations.
I think the issue is that the higher RAM offers tend to be higher disk so they don't end up in that price range as frequently. I'll see if I can tune things to where it's more plausible for that to exist in that "sweet spot" as I bet a lot of other people would also be interested.
Like that $33-45 offer in San Jose recently could have been $20-25 if it wasn't with 180GB disk and it was 14TB data, 3.8GB memory, 3 cores, and then multiple others that would have hit that target if they weren't 200-250GB disk.
I'm actually going to poke around in the generator a bit right now, and let's see what it spits out instead. I was too late for the generations that were coming right up so it'll be like a few hours later on the next batch. Let me know if it improves.
Not sure if this is really relevant for this thread, but I had some deployment troubles on "Node RYZE.AMS-D001.VMS". The reinstall for "[Ryzen Compatible] Ubuntu 24 (64 Bit)" in full control panel claimed to work, but didn't. After failing to log in, changing the root password resulted in logging in to Alma. The "[Ryzen Compatible] Ubuntu 20.04 (64-Bit)" did work, but having to change the "ru" apt sources is a bit of unnecessary work (and the plethora of updates to get to dist upgrade stage).
The Ubuntu 24 on "Node RYZE.AMS-D026.VMS" I installed (a day or so earlier) on the other hand did work.
The 20.04 is nearing EOL, so having something newer as an install option (without the full control panel) might be nice at some point.
On slightly more unrelated note, I also reset my idling surprise box, and since 384MB boxes seem to be on sale, some recommendations on distros based on the memory amount might be a nice feature. I tried Ubuntu minimal (did not manage to log in), Debian 11 (Segmentation fault at apt upgrade*), and then got Debian 10 to work (even apt).
[ 126.501227] apt-listchanges[844]: segfault at 182d0d2b8 ip 00000000004fd56c sp 00007ffc47753fd0 error 4 in python3.11[41f000+2b3000] likely on CPU 0 (core 0, socket 0)
And while the above rambling may sound a bit negative, I'm actually quite happy with the vms (and VirMach). Just noting that there feels to be a bit of work that feels a bit unnecessary (at least with Ubuntu).
@VirMach said: Basic details now appear on your services page. Each of them has a manage button the same way it would on the combined SolusVM account. And if you open the control panel on an account that only has one virtual server, it auto opens the management page for that one specifically.
Also I think I made it to where it by default filters to active services only on that page to clean it up without you having to click "Active" on the sidebar. There were also some other controls but they were a little buggy so I only kept what works.
Update to this new feature. I figured it wouldn't be very useful if you had to wait 5 to 10 seconds plus for them to load to see the details and be able to click the control panel button. I've removed the (1) bandwidth usage and amount, (2) server status, (3) node name. This has improved the load times tremendously. I may add back the code for them but have it load independently from the rest of the stuff so you can quickly see the memory, disk, CPU, service IP address, as well as control panel and VNC buttons, and then if you still have the page open it will grab the other stuff afterward.
Actually I went ahead and did that so that's how it'll work now, better. Also noticed a bug on the part of the code that auto selects "Active" services on page load, I'll work on that sometime later.
@wintah said:
Not sure if this is really relevant for this thread, but I had some deployment troubles on "Node RYZE.AMS-D001.VMS". The reinstall for "[Ryzen Compatible] Ubuntu 24 (64 Bit)" in full control panel claimed to work, but didn't. After failing to log in, changing the root password resulted in logging in to Alma. The "[Ryzen Compatible] Ubuntu 20.04 (64-Bit)" did work, but having to change the "ru" apt sources is a bit of unnecessary work (and the plethora of updates to get to dist upgrade stage).
The Ubuntu 24 on "Node RYZE.AMS-D026.VMS" I installed (a day or so earlier) on the other hand did work.
The 20.04 is nearing EOL, so having something newer as an install option (without the full control panel) might be nice at some point.
On slightly more unrelated note, I also reset my idling surprise box, and since 384MB boxes seem to be on sale, some recommendations on distros based on the memory amount might be a nice feature. I tried Ubuntu minimal (did not manage to log in), Debian 11 (Segmentation fault at apt upgrade*), and then got Debian 10 to work (even apt).
[ 126.501227] apt-listchanges[844]: segfault at 182d0d2b8 ip 00000000004fd56c sp 00007ffc47753fd0 error 4 in python3.11[41f000+2b3000] likely on CPU 0 (core 0, socket 0)
And while the above rambling may sound a bit negative, I'm actually quite happy with the vms (and VirMach). Just noting that there feels to be a bit of work that feels a bit unnecessary (at least with Ubuntu).
I wrote out a long reply to this but then realized I'm just being confusing and not very helpful. If you found any specific issues that end up preventing you from being able to do what you want, make a ticket and let us know it was pre-approved and your username here (since beta tickets will probably end up closed otherwise) and I'll have a look.
@nullnothere said: It's mostly looking at balanced offers from the VirBot vs other BF offers around. Something like a 2-4 vCPU/4GB RAM/60GB Disk/5TB BW at roughly 20-25 USD (or EUR, stretch a bit) equivalent is very reasonable and attractive and I'd even hazard reasonably "frequent" during BF across locations. Add a no-ipv6 penalty/discount (for those of us who do care about IPv6) and
Okay, I'm seeing the plans it generated and they definitely look "better." It could bee luck of the draw. None of them are what you want but some are getting closer. I'll do one as a direct comparison since it re-rolled it from a previous exact server and also happened to get the same exact randomizer roll (meaning it stays kind of similar.)
So 14% lower price, 28% more memory, 24% less disk, but 50% less bandwidth (the bandwidth is kind of RNG.)
Napkin math, that means something like 30-50% more RAM for the price while keeping the actual pricing formulas kind of similar. It just does some other things (which can indirectly reduce price for a specific resource and higher chance of getting more.)
Not sure if this is beta, but it took payment and then redirected me to an unpaid invoice page. Not going to setup a ticket as we're in the beta stages. Love the flash deal idea page.
EDIT: I'm dumb, it took a pre-auth but returned it. Time to re-go through checkout. Probably didn't like address.
Comments
The scrolling looks weird.
https://imgur.com/nAmZRrd
There's no way I'm touching that. I'm not very good with trying to get it to work in the first place so it's a miracle. I think I spent probably an hour on that and it was bouncing all over the place. This is as close to normal as I can get it unless someone wants to volunteer an improvement.
I'm guessing that's on a specific resolution I didn't test. Oh yeah it's if the screen is smaller, replicated. Get a bigger monitor and fullscreen it so I don't have to fix it. Or make it even smaller.
Oh it's based on height size not width. Let me see... okay fixed by disabling it based on height as well. That was easy.
Order in AMS pending since 24h
I guess it’s normal and I’m not VIRMARCHED
Some people talked about notifications earlier and I had something like this written out but I'm trying to compile a list of improvements wee should make while keeping them realistic.
I'll also try to also add in the plan number as requested for the backend notifications. I need more information on it being late or "phantom" because it matters if it's the backend one, frontend, etc. Also some of it out of our control. It depends on your browser, your operating system and how they both manage notifications. They've made a lot of "smart" changes where they hide them or delay them based on your patterns and I don't believe we can override that. I might also add a timestamp, probably in UTC, for the backend ones as those seem to be delayed like that more than the frontend.
One of the "smart" features I think we could take advantage of "taking back" thee previous notification and/or updating it. I believe the way we have it set up is a unique notification per plan ID, but we could possibly clean it up to where it goes based off each product card, if I can figure that out. Then that way it should get updated and you won't end up with like 20 notifications on your notification sidebar.
I'm still not done making a complete mess on the backend. Give me your order information so I can make sure I didn't miss yours specifically when cleaning the errored ones manually.
What happens after the heat death of the universe, will my discount still be recurring?
Order #611314 It's almost 24 hours. It still hasn't turned on.
Any other feedback or suggestions or requests you have on how things should work, speak now so we have time to include it in potential changes.
This is because of an error catcher, if it catches an error in configuration settings it rolls it back and doesn't activate it. Before it would activate it with the wrong specifications and I'd manually fix it. Once the code's fixed it should generate. I might remove the validator soon so they can just generate (wrong) and I can go through and fix them versus just being stuck. I'll decide in a few hours.
Invoice 1661927
Now marked as active but "Virtual server not found" appears in the panel
EDIT: Can be installed from solusvm but not from customer panel (Virtual server not found)
Yeah same thing, this is actually not a vserverid error so it's fine. It's just marked as active but not actually active. I might add for it to go back to pending status so people don't think it's broken but it's better to just fix it so it validates.
I fixed the logic, but the rollback logic had an error. So yours actually just created but rolled back. I don't know what happens now, definitely don't use it on SolusVM right now, it might get terminated.
Either it will re-create now and old service will stay there and manually cleaned up/terminated, or it'll get permanently stuck until I manually go through these that created a few minutes ago and fix them.
Ok no problem. I'll just wait your fixes
I was deployed on the same node as my previous VM, no luck
Should have finally provisioned correctly. Well almost, I'm updating it manually to the final product. Looks good now.
Just another reminder on duplicate services - don't use the one that's not on WHMCS. I mean you could use it, I don't care, but the duplicates during the BETA will likely have NO grace period and NO support in switching back to it, recovering data, etc. If we forget to clean them up, I don't want a ticket several months from now asking to grab data off the one that wasn't on WHMCS.
Oh wait I forgot I have a tool for this, I'm going to run it now for the newly created ones so we can eliminate that confusing possibility.
I've ran it now, the one(s) you shouldn't use are suspended and the correct one is active. Let me know if anything else looks weird since yours got created through like 3 or 4 code changes. Basically report anything weird immediately and definitely keep backups for any super important files around 2 weeks per the recommendation (or honestly, you should always do that anyway.)
You have test ip each location like Amsterdam, San jose, Atlanta and Chicago? Looking glass seems out-dated.
It's on my list to fix, they're slowly being updated. Hopefully by tomorrow?
But here's some IPs that may or may not work:
SJC - 91.186.208.96, 83.147.241.226
CHI - 85.209.17.230
ATL - 82.153.68.71
AMS - 88.214.23.14, 78.142.228.4
It'll be nice to have 3 different locations to pick from (along with the different specs/prices of course).
Also, for the crazy people here (myself included), since you've risen from the ashes (yet again?), can you:
a) Reward those of us who have renewals (and still have faith in you for whatever weird reason, despite not even knowing where the instance is running or if it is in existence at all - does a VPS that doesn't answer pings exist at all)?
b) Some recurring coupons please? So far everything that I've seen is a one off (very generous at that of course) but it'll be nice to encourage some longer term customers who would like to stick with you (or should I say tolerate) for a year or two more, especially if the price is attractive?
c) Can you throw the old faithful a few hidden gems and treasures for a seek-and-you-shall-find kind of a game? You know we're suckers for hard to get things at low prices even if they are not going to be used beyond the trophy-I-snagged-it stage.
d) How about a nice "bundle" (or a pick any 3-5 locations type of thingy) - same (decent low end specs 512MB or 1GB of RAM, 1 or 2 vCPUs, 2-5 TB of BW, 25-40GB disk, 1 IPv4 + IPv6... yeah I know you don't do IPv6 but come on... it's 2024 and you've been promising for a few years now and finally you're out of clutches of that DC that doesn't do IPv6), all locations
OK. I've run out of ideas other than creating a cheeky VeerBot that veers off the specs every now and then and shocks you with some insane offer.
If you've read this far, surely it was entertaining enough and you should throw us an offer that we can't refuse.
Thank you...
Let's just say it might or might not have already happened...
I think if we did any work to get this done, it might actually end up being worse than how it's done now. The current set of offers that are about to go live (they just recently expired as I'm writing this) are Tokyo, Amsterdam, and San Jose. If you look at every chunk of 3, there's usually at least two different locations. If we force it to keep plans in different locations, some locations would end up running out sooner and making the future diversity worse, and the plan diversity would also suffer. I feel like randomized is the most interesting pattern.
Those who stick around are naturally already skewed toward cheaper services, so it's difficult to provide any type of discount on top of that. Also your statement, while true in some very specific clusters/cases, is not anywhere close to the actual big picture. You're focusing on let's say 5% of all cases. The other 95% is mostly several hundred days of uptime. We've also already been (usually) generous with those major outages/issues with service extensions. If you point out specific cases, I can take a further look and make sure that people were appropriately credited.
I know there's a psychological thrill to having prices be higher and then receiving a discount on top of it, but take a look at the specials as they are currently. They already get further price cuts automatically in most cases if they hang around too long, and on the lower end you can easily tell that the price formulas are set up in a way where they barely even cover IPv4 costs. There's a calculated percentage off on all of them, and that's pretty accurate compared to our regular plans, with most of them being in the 80% or higher discount range, some hitting like 88% or more. So with that said, I'm unsure how much more you want off recurring to consider it to be "attractive."
Perhaps I haven't looked around much recently. I know you're setting up a lot of this as "VirMach's intolerable" so my next question might have the answer that "VirMach is intolerable so it has to be cheaper"... but are there any other providers with active offers, on newer hardware, that are providing deals with an IPv4 for lower than $6 to $8 a year?
By the way the idea behind the one time discount isn't to try to encourage more sales, it's just a way to give people the first few months free because something might go wrong and the service might get stuck/unusable for a while. That hasn't really happened but it was just a precaution.
We're barely back to being able to do a sale, and even now, there's other outstanding issues and not a lot of time. I wish I could, it's fun for me as well, but the second issue is that as I mentioned earlier, renewals are skewed toward these harder finds and over the last 5-10 years, the numbers of them have gone up, and they're now a higher percentage of all total services. So we have to be careful about the quantity. I'm not saying there won't be any, but the quantity that would be available would be so rare that it's likely not worth the time & effort for anyone this time around.
This might be possible but it wouldn't be a bundle with flash sales. It'd have to be some other set bundle.
I really do want to get IPv6 going but SolusVM support for it is abysmal. I don't feel comfortable doing it and if we did it I would want to not introduce more problems. We just haven't had the time to finish the system required to do this with confidence. I don't want to give any timelines because as you said, it's been a while. It'll happen eventually.
And yes it's kind of ironic how much we even personally made fun of that DC that doesn't do IPv6 and we kind of ended up turning into them.
Oh I didn't realize that I was that far out...
Anyway I hope there's redemption yet again is an option.
I guess, I have let my guard down thinking it's too early but... en-guarde!
I am blind. Didn't read your wallll of text for deal numbers in notifications.
Oh. I didn't think this through from your perspective (esp. varying stock levels vs popularity). Fair enough.
I'd like to think so, but I have also ended up pruning my "collection" to a bare minimum because I lost track of where what was and if it was running at all (through all the migrations and what not). So I'm not very clear on what happened (or why). I'm now going to look at a few low end offers via the VirBot and maybe pick a few here and there just for kicks. I've mostly abandoned hopes of a big powerful VM somewhere because of the huge migrate-away-from-the-evil-DC fiasco whenever it happened (see - I've lost track of time as well.
It's mostly looking at balanced offers from the VirBot vs other BF offers around. Something like a 2-4 vCPU/4GB RAM/60GB Disk/5TB BW at roughly 20-25 USD (or EUR, stretch a bit) equivalent is very reasonable and attractive and I'd even hazard reasonably "frequent" during BF across locations. Add a no-ipv6 penalty/discount (for those of us who do care about IPv6) and
I think you're entering pretty decent sale territory.
Surprisingly there are a couple who are in that price range (and already have sold out) - but let's be realistic, no one is looking for a mass deadpool offer. A few sparse crazy offers to up the ante and the regular balanced sale offers (hopefully sustainable) will be what people are looking for mostly.
That's fair (and appropriate) - I didn't think that way at all (see, I'm thinking only "consumer" side not "producer" side)
Agreed and again, not a perspective I had considered. I do understand on the outstanding issues which we do want fixed (and that is why I think this thread is popular - you're responsive or at least good at faking it!) - and I'm also quite happily engaged because you did take the time to put together a rather long, well thought out, point by point justification of your stance (note I didn't use the word rebuttal here!).
Well... how come so many of the other providers (many of whom can't even get emails to work properly) are able to provide good IPv6 functionality with or without SolusVM? You're literally the ONLY provider I have VMs with that don't have a proper functional IPv6 setup (at most locations) but but but because it's VirMach, I'm kind of still OK with a few IPv4 only test vms (again, because it's cheap enough, mostly at the <$10 price point, that I don't mind "splurging"). But it's getting to be an irritant and I really hope you can at least make some sort of a fix for the crazy knowledgeable ones to use IPv6.
Thank you for taking the time to go though and put together your points.
Now where is that reward for such a flawless work of literature on the merits of engaging nicely with a provider to ask for a shameless discount coupon that lasts to eternity and beyond?
I think the issue is that the higher RAM offers tend to be higher disk so they don't end up in that price range as frequently. I'll see if I can tune things to where it's more plausible for that to exist in that "sweet spot" as I bet a lot of other people would also be interested.
Like that $33-45 offer in San Jose recently could have been $20-25 if it wasn't with 180GB disk and it was 14TB data, 3.8GB memory, 3 cores, and then multiple others that would have hit that target if they weren't 200-250GB disk.
I'm actually going to poke around in the generator a bit right now, and let's see what it spits out instead. I was too late for the generations that were coming right up so it'll be like a few hours later on the next batch. Let me know if it improves.
Not sure if this is really relevant for this thread, but I had some deployment troubles on "Node RYZE.AMS-D001.VMS". The reinstall for "[Ryzen Compatible] Ubuntu 24 (64 Bit)" in full control panel claimed to work, but didn't. After failing to log in, changing the root password resulted in logging in to Alma. The "[Ryzen Compatible] Ubuntu 20.04 (64-Bit)" did work, but having to change the "ru" apt sources is a bit of unnecessary work (and the plethora of updates to get to dist upgrade stage).
The Ubuntu 24 on "Node RYZE.AMS-D026.VMS" I installed (a day or so earlier) on the other hand did work.
The 20.04 is nearing EOL, so having something newer as an install option (without the full control panel) might be nice at some point.
On slightly more unrelated note, I also reset my idling surprise box, and since 384MB boxes seem to be on sale, some recommendations on distros based on the memory amount might be a nice feature. I tried Ubuntu minimal (did not manage to log in), Debian 11 (Segmentation fault at apt upgrade*), and then got Debian 10 to work (even apt).
And while the above rambling may sound a bit negative, I'm actually quite happy with the vms (and VirMach). Just noting that there feels to be a bit of work that feels a bit unnecessary (at least with Ubuntu).
Update to this new feature. I figured it wouldn't be very useful if you had to wait 5 to 10 seconds plus for them to load to see the details and be able to click the control panel button. I've removed the (1) bandwidth usage and amount, (2) server status, (3) node name. This has improved the load times tremendously. I may add back the code for them but have it load independently from the rest of the stuff so you can quickly see the memory, disk, CPU, service IP address, as well as control panel and VNC buttons, and then if you still have the page open it will grab the other stuff afterward.
Actually I went ahead and did that so that's how it'll work now, better. Also noticed a bug on the part of the code that auto selects "Active" services on page load, I'll work on that sometime later.
I wrote out a long reply to this but then realized I'm just being confusing and not very helpful. If you found any specific issues that end up preventing you from being able to do what you want, make a ticket and let us know it was pre-approved and your username here (since beta tickets will probably end up closed otherwise) and I'll have a look.
Okay, I'm seeing the plans it generated and they definitely look "better." It could bee luck of the draw. None of them are what you want but some are getting closer. I'll do one as a direct comparison since it re-rolled it from a previous exact server and also happened to get the same exact randomizer roll (meaning it stays kind of similar.)
How it looked before update:
How it looks after:
So 14% lower price, 28% more memory, 24% less disk, but 50% less bandwidth (the bandwidth is kind of RNG.)
Napkin math, that means something like 30-50% more RAM for the price while keeping the actual pricing formulas kind of similar. It just does some other things (which can indirectly reduce price for a specific resource and higher chance of getting more.)
Not sure if this is beta, but it took payment and then redirected me to an unpaid invoice page. Not going to setup a ticket as we're in the beta stages. Love the flash deal idea page.
EDIT: I'm dumb, it took a pre-auth but returned it. Time to re-go through checkout. Probably didn't like address.