@VirMach said: and most people are not just using it for storing personal files/backups.
Yo chad, them have been given too much bandwidth limit.
6TB traffic per month for a 500GB storage service, just encourage folks use it as seed-box, website static file serve, whatever.
Worth to mention that your counterpart GreencloudVPS which in the same data-center, only give 1TB per month for the same disk size.
I agree. bandwidth should be double, no more than triple the disk size otherwise people will abuse it.
Does torrenting consume a lot of CPU power though?
Can you limit the CPU usage for all storage VPS users? Like a user can only burst up to 100% of one core per 500GB storage for no longer than half an hour every 24 hours, as other providers greencloudVPS included did. @VirMach
@Daevien said:
Hmm 6tb for 500g is more than my storage servers have yeah. I wouldn't mind having more bandwidth on some of my larger vps on Virmach heh. And adjusting down any packages now would be troublesome. The fact that it's new ryzen cores even if it's just one core means it's potentially a lot of power. And I'm not sure the limiters in SolusVM work all that well. No big surprise, Cpanel and SolusVM are brothers in terribleness and broken "features".
For comparison, my three vps storage servers are (others are local or stored with friends):
InceptionHosting - 1x 2200mhz core, 1 tb space, 1tb bandwidth, special deal i've had since Aug 2020
Letbox - 1x ryzen 3900 core, 1tb space, 10tb bandwidth, older deal marked as legacy from june 2021
Wishosting - 1tb space with 250mbit unmetered and no real ip, just some ports behind NAT (cores / cpu / location have all changed due to length of time (jan 2018) and migrations of the package and some downtime so it's hard to compare those heh)
So I guess yeah that much transfer that close to home made them really pay attention and go crazy to get them.
@VirMach said: and most people are not just using it for storing personal files/backups.
Yo chad, them have been given too much bandwidth limit.
6TB traffic per month for a 500GB storage service, just encourage folks use it as seed-box, website static file serve, whatever.
Worth to mention that your counterpart GreencloudVPS which in the same data-center, only give 1TB per month for the same disk size.
This is a valid point but there are definitely use cases that are able to use all the bandwidth and not cause problems. Also, the reason bandwidth is so high is because we basically kept it the same, even for Tokyo, and the reason why others in the same DC and probably same network give 6x less is because bandwidth there is insanely expensive probably to the tune of 6x and we just decided to eat the cost there, so yes it encourages some bad behavior.
If it was 100% true though and there wasn't also other differences, then LAXA004S should be having the same/similar problems. But it's definitely just different regional use cases and peaks, perhaps somewhat also affected by the fact that it's the only affordable option in the region.
@foitin said:
Can you limit the CPU usage for all storage VPS users? Like a user can only burst up to 100% of one core per 500GB storage for no longer than half an hour every 24 hours, as other providers greencloudVPS included did. @VirMach
Do they do this in any particular automated manner, without a reboot?
@Daevien said:
Did you try vnc to see if it actually boots up?
Yes, but the host is marked as offline and has been for awhile. I've tried powering off, re-installing, rescue, and probably other things without any luck. I've seen the eth0->ens3 change that often needs to be done, but this does not appear to be the issue.
@VirMach said:
I have no idea how they haven't fixed that or at least allow people to click a button client-side but what you could do is disable VNC to see if it comes online I think and that confirms that's the issue.
I tried disabling that, powering off, then booting up, still offline. That doesn't seem to have made any difference. If you want me to open a ticket let me know, but I assume this thing gets no free support. It's a $5/year BF special.
@foitin said: I agree. bandwidth should be double, no more than triple the disk size otherwise people will abuse it.
Does torrenting consume a lot of CPU power though?
We don't sell seedboxes for a reason, we don't know that much about it and haven't ever investigated it thoroughly but I do have some personal experience. Torrenting is heavily dependent, as far as I can tell, on memory for it not to absolutely thrash I/O. We did increase memory amounts but all that did is probably encourage it, but it's probably less than optimal on the smaller packages and it's not like people are torrenting like a normal user, perhaps some are, but in terms of high usage, they're doing mass quantities at a time and that matters as well. After all a normal user doesn't need like 2TB just for a regular personal amount of usage. It's combined with so many other things and clearly shared by multiple people in those cases where the bandwidth usage is high.
Again though I have to re-iterate, that's only one of two types of problems. We essentially already took care of that to a good degree, it's the second type of problem that's difficult to address.
@jtk said: Yes, but the host is marked as offline and has been for awhile. I've tried powering off, re-installing, rescue, and probably other things without any luck. I've seen the eth0->ens3 change that often needs to be done, but this does not appear to be the issue.
I'm a little confused, if it's marked as offline, what do you mean regarding the eth0-->ens3 change? It'd have to be powered on to modify network configuration within the OS.
I'm going to just assume you're saying it does actually boot up then, into OS? In that case the VNC thing is irrelevant if it also boots up with it on. Yeah you can make a ticket, no problem, just please make sure it goes into the right department.
@VirMach said:
I'm a little confused, if it's marked as offline, what do you mean regarding the eth0-->ens3 change? It'd have to be powered on to modify network configuration within the OS.
That was in response to another commenter's suggestion. I know that sometimes a problem arises that the interface name has changed with a kernel update, and that networking will not be operational until that is changed in the config. That is simply not the case here.
I'm going to just assume you're saying it does actually boot up then, into OS? In that case the VNC thing is irrelevant if it also boots up with it on. Yeah you can make a ticket, no problem, just please make sure it goes into the right department.
mysqlreplication_FFME004 is now UP.
Downtime: 2d 34 min
It's aaaaaaaaaaaaaaaaaalive.
Good thing it's NVMe and syncing up 2 days of mysql replication won't get my suspended for abuse (I hope, it's running for an hour already) Seconds_Behind_Master: 62802
17 more hours to go, seems like it done 1 day of data within an hour, so I should stop abusing within next hour when everything catches up!
Oh btw. @VirMach any interesting story with FFME004? Mobo? CPU? Cables? No-idea, we just asked DC hands to re-seat stuff?
@Jab said:
Oh btw. @VirMach any interesting story with FFME004? Mobo? CPU? Cables? No-idea, we just asked DC hands to re-seat stuff?
About 30 reboots and now monitoring it 24/7 to check for the first sign of disk deciding to do the thing it did again while planning potential emergency migrations for people on that disk. Fun stuff.
@VirMach I know things are busy, but I'm still waiting on that OS change for our dedi server. Ticket #218112.
Trying to be patient, but really would like to get this going and it's been a couple months since it all started.
Was taking a look at your new specials preview and I noticed that everything looks really well scaled except for the 4096 special.
As an example, combining the 512 and 1024 specials ($22+$33=$55) gives you greater/equal specs of the two combined.
However, combining the 1024 and 2048 specials ($33+$55=$88) gives you 4TB bandwidth instead of at least 5TB.
I realize that prices are arbitrary, but it would be nice to see this weird combination scaling work out considering that the price points seem to line up as well.
@Civalo said: @VirMach I know things are busy, but I'm still waiting on that OS change for our dedi server. Ticket #218112.
Trying to be patient, but really would like to get this going and it's been a couple months since it all started.
Thanks,
John
Sorry, it shouldn't have taken this long. I do see why but we should have identified it and changed the ticket title accordingly. We usually received "Dedicated Server Migration" tickets per the email we sent out and if conversation changed to OS re-install they're processed on the same ticket. I'll get it corrected, and credit you for the 2~ months (plus the initial month so actually 3.)
(edit) Basically since it was called "Change of OS" it unfortunately didn't get flagged. We did do a search query for "dedicated" and renamed all those but "dedicated" wasn't mentioned in the message body either. I'm not trying to blame you or anything like that, just explaining where we failed to pick this one up and put it in the same queue. Otherwise it essentially looked like a lot of other tickets we have a backlog on that are considered not priority since we assume it's customer requesting OS re-install on a VPS which was already corrected around that time for some servers.
@ben said:
Boss, now the function to increase funds is not working, node: Phoenix
Same here.
Seems that affected a bunch of people.
Maybe, it's closed on purpurse.
Oh, VirMach checked into my account and said that it was a bug and they'd have a dev address it, so I doubt it's on purpose. I'm sure getting paid is fairly high on the priority list so I'd bet it won't stay bugged for long.
I've only used it for a while. Probably know that running full network speed will immediately disconnect. There were a few other times when I used full cpu and also could not connect to ssh, not sure if it was disconnected or offline.
@foitin said:
Can you limit the CPU usage for all storage VPS users? Like a user can only burst up to 100% of one core per 500GB storage for no longer than half an hour every 24 hours, as other providers greencloudVPS included did. @VirMach
Do they do this in any particular automated manner, without a reboot?
not sure. never run intensive tasks on it.
but it's clearly stipulate in their terms.
from what i heard they shut down cpu abusers as a warning first.
This is a valid point but there are definitely use cases that are able to use all the bandwidth and not cause problems. Also, the reason bandwidth is so high is because we basically kept it the same, even for Tokyo, and the reason why others in the same DC and probably same network give 6x less is because bandwidth there is insanely expensive probably to the tune of 6x and we just decided to eat the cost there, so yes it encourages some bad behavior.
Then you shouldn't eat it. Just lower the bandwidth to 1/6 of current amount.
No extra cost and better user experience for legit users. Win-win.
Is this the reason you put 10G port on hold? Too many abusers?
@foitin said: Then you shouldn't eat it. Just lower the bandwidth to 1/6 of current amount.No extra cost and better user experience for legit users. Win-win.Is this the reason you put 10G port on hold? Too many abusers?
I'm not sure how virmach thinks anymore.Suspended machines that I never used are.It's not clear if he just wants to suspended Tokyo users now or to handle abusers?
Hope he will have more time to think and solve the existing problems.
Comments
This is a valid point but there are definitely use cases that are able to use all the bandwidth and not cause problems. Also, the reason bandwidth is so high is because we basically kept it the same, even for Tokyo, and the reason why others in the same DC and probably same network give 6x less is because bandwidth there is insanely expensive probably to the tune of 6x and we just decided to eat the cost there, so yes it encourages some bad behavior.
If it was 100% true though and there wasn't also other differences, then LAXA004S should be having the same/similar problems. But it's definitely just different regional use cases and peaks, perhaps somewhat also affected by the fact that it's the only affordable option in the region.
Do they do this in any particular automated manner, without a reboot?
Yes, but the host is marked as offline and has been for awhile. I've tried powering off, re-installing, rescue, and probably other things without any luck. I've seen the eth0->ens3 change that often needs to be done, but this does not appear to be the issue.
I tried disabling that, powering off, then booting up, still offline. That doesn't seem to have made any difference. If you want me to open a ticket let me know, but I assume this thing gets no free support. It's a $5/year BF special.
Dataplane.org's current server hosting provider list
We don't sell seedboxes for a reason, we don't know that much about it and haven't ever investigated it thoroughly but I do have some personal experience. Torrenting is heavily dependent, as far as I can tell, on memory for it not to absolutely thrash I/O. We did increase memory amounts but all that did is probably encourage it, but it's probably less than optimal on the smaller packages and it's not like people are torrenting like a normal user, perhaps some are, but in terms of high usage, they're doing mass quantities at a time and that matters as well. After all a normal user doesn't need like 2TB just for a regular personal amount of usage. It's combined with so many other things and clearly shared by multiple people in those cases where the bandwidth usage is high.
Again though I have to re-iterate, that's only one of two types of problems. We essentially already took care of that to a good degree, it's the second type of problem that's difficult to address.
I'm a little confused, if it's marked as offline, what do you mean regarding the eth0-->ens3 change? It'd have to be powered on to modify network configuration within the OS.
I'm going to just assume you're saying it does actually boot up then, into OS? In that case the VNC thing is irrelevant if it also boots up with it on. Yeah you can make a ticket, no problem, just please make sure it goes into the right department.
Once hosthatch's VPS users use cpu 100% for a few minutes, their machines are automatically offline for ten minutes.
That was in response to another commenter's suggestion. I know that sometimes a problem arises that the interface name has changed with a kernel update, and that networking will not be operational until that is changed in the config. That is simply not the case here.
It does not boot up. I'll open a ticket.
Dataplane.org's current server hosting provider list
Yep, it is typical for BuyVM. They will cap CPU usage.
Also Contabo will cap disk IOPS.
Ontario Dildo Inspector
mysqlreplication_FFME004 is now UP.
Downtime: 2d 34 min
It's aaaaaaaaaaaaaaaaaalive.
Good thing it's NVMe and syncing up 2 days of mysql replication won't get my suspended for abuse (I hope, it's running for an hour already)
Seconds_Behind_Master: 6280217 more hours to go, seems like it done 1 day of data within an hour, so I should stop abusing within next hour when everything catches up!
Oh btw. @VirMach any interesting story with FFME004? Mobo? CPU? Cables? No-idea, we just asked DC hands to re-seat stuff?
Haven't bought a single service in VirMach Great Ryzen 2022 - 2023 Flash Sale.
https://lowendspirit.com/uploads/editor/gi/ippw0lcmqowk.png
About 30 reboots and now monitoring it 24/7 to check for the first sign of disk deciding to do the thing it did again while planning potential emergency migrations for people on that disk. Fun stuff.
@VirMach I know things are busy, but I'm still waiting on that OS change for our dedi server. Ticket #218112.
Trying to be patient, but really would like to get this going and it's been a couple months since it all started.
Thanks,
John
Was taking a look at your new specials preview and I noticed that everything looks really well scaled except for the 4096 special.
As an example, combining the 512 and 1024 specials ($22+$33=$55) gives you greater/equal specs of the two combined.
However, combining the 1024 and 2048 specials ($33+$55=$88) gives you 4TB bandwidth instead of at least 5TB.
I realize that prices are arbitrary, but it would be nice to see this weird combination scaling work out considering that the price points seem to line up as well.
Sorry, it shouldn't have taken this long. I do see why but we should have identified it and changed the ticket title accordingly. We usually received "Dedicated Server Migration" tickets per the email we sent out and if conversation changed to OS re-install they're processed on the same ticket. I'll get it corrected, and credit you for the 2~ months (plus the initial month so actually 3.)
(edit) Basically since it was called "Change of OS" it unfortunately didn't get flagged. We did do a search query for "dedicated" and renamed all those but "dedicated" wasn't mentioned in the message body either. I'm not trying to blame you or anything like that, just explaining where we failed to pick this one up and put it in the same queue. Otherwise it essentially looked like a lot of other tickets we have a backlog on that are considered not priority since we assume it's customer requesting OS re-install on a VPS which was already corrected around that time for some servers.
Boss, now the function to increase funds is not working, node: Phoenix
sorry OOT.
@Franzkafka
Hey bro, Just saw your name as one of the contributors of various proxy softwares (shadowsocks,xray, etc).
Are you the same person?
https://microlxc.net/
Same here.
Seems that affected a bunch of people.
Maybe, it's closed on purpurse.
Oh, VirMach checked into my account and said that it was a bug and they'd have a dev address it, so I doubt it's on purpose. I'm sure getting paid is fairly high on the priority list so I'd bet it won't stay bugged for long.
Whaaaaagh, could you cut the cost of change GFW blocked IP to 3$?
Btw, why Tokyo storage VPS can not purchase additional IP?
Ontario Dildo Inspector
My guess would be because a storage vps shouldn't need another IP, it's for STORAGE. Not webhosting / email / etc
Whatever...
Maybe it still need some efforts to be "shutup and take my money"
I simple want to buy another japan VPS as jump proxy. Wtf is Japan setup & unlock 80/443 fee. Whaaaaaaaaaaaaaaaaaaaaagh!!!
It certainly will stimulate account trading/renting.
Ontario Dildo Inspector
lol ok
I bench YABS 24/7/365 unless it's a leap year.
No. Go spread misinformation elsewhere.
I've only used it for a while. Probably know that running full network speed will immediately disconnect. There were a few other times when I used full cpu and also could not connect to ssh, not sure if it was disconnected or offline.
Surprised to see you here. Can I ask if ipv6 at tokyo location stopped working for everyone or just me?
not sure. never run intensive tasks on it.
but it's clearly stipulate in their terms.
from what i heard they shut down cpu abusers as a warning first.
not just you. it's been down for what? like around 3 months already?
and funny thing he's doing ego searching from time to time.
otherwise he wouldn't notice it right? his last comment is more than 3 months ago.
Then you shouldn't eat it. Just lower the bandwidth to 1/6 of current amount.
No extra cost and better user experience for legit users. Win-win.
Is this the reason you put 10G port on hold? Too many abusers?
I'm not sure how virmach thinks anymore.Suspended machines that I never used are.It's not clear if he just wants to suspended Tokyo users now or to handle abusers?
Hope he will have more time to think and solve the existing problems.