Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


BIG DISK OFFERS: $2.4/TB all-in storage VPS and Ryzen 5950X w/big block storage! - Page 7
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.

BIG DISK OFFERS: $2.4/TB all-in storage VPS and Ryzen 5950X w/big block storage!

145791012

Comments

  • yongsikleeyongsiklee Member, Patron Provider

    ... We're exploring (based on some feedback and requests) an NVMe cached/accelerated HDD storage solution currently. We are also contemplating all-NVMe big storage plans--though they would be more expensive than HDD. I'm not sure what kind of appetite exists for those since they'll likely be roughly around the $11 per TiB mark.

    All NVMe storage? Do you see any demand for it in the near future? Otherwise, forget it unless you hit a jackpot.

    Thanked by 1greentea
  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @namein said:

    @crunchbits said: roughly around the $11 per TiB mark.

    the price is for per month? That was indeed a bit beyond expectations.

    I've had a great time with Crunchbits so far, and I'm impressed with the well-designed admin panel. Their technical support response is also very quick, and I think I'll be switching to an annual contract with them soon for the long haul.

    @yongsiklee said:

    ... We're exploring (based on some feedback and requests) an NVMe cached/accelerated HDD storage solution currently. We are also contemplating all-NVMe big storage plans--though they would be more expensive than HDD. I'm not sure what kind of appetite exists for those since they'll likely be roughly around the $11 per TiB mark.

    All NVMe storage? Do you see any demand for it in the near future? Otherwise, forget it unless you hit a jackpot.

    This wouldn't be a low-end focused offer. We do have users that require bigger storage (generally complimenting a more powerful VPS/Dedi) with speed requirements to match, but as far as I can tell SSDs are still plenty fast for their uses. The only difference is we actually get NVMe storage for less than SSDs now.

    This has been tied mostly to our GPU/HPC users. I guess sometimes(often?) their workloads outpace the ~140MB/s of spinning disk arrays.

  • Hey guys, can you tell me which city the VPS's are physically in? My VPS IP resolves to Tampa Florida. Just interested in terms of distance/latency from me in NZ.

    Thanked by 1greentea
  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @andynz said:
    Hey guys, can you tell me which city the VPS's are physically in? My VPS IP resolves to Tampa Florida. Just interested in terms of distance/latency from me in NZ.

    Some of our newer IPv4 ranges show Tampa. They're not in Tampa, or anywhere near Florida. It's a geodb not-yet-updated thing it seems, if you check other sources it will show correctly (Cheney, WA or CDA, ID) but it's about 50/50.

    Thanked by 1greentea
  • any chance of a restock of 4TB Storage VPS?

  • @crunchbits said:

    @andynz said:
    Hey guys, can you tell me which city the VPS's are physically in? My VPS IP resolves to Tampa Florida. Just interested in terms of distance/latency from me in NZ.

    Some of our newer IPv4 ranges show Tampa. They're not in Tampa, or anywhere near Florida. It's a geodb not-yet-updated thing it seems, if you check other sources it will show correctly (Cheney, WA or CDA, ID) but it's about 50/50.

    Great thanks. Explains similar latency to Seattle services I have.

    To labour a point :), just amazing service guys and great full featured control panel. Can the storage size be increased at a future date without losing current data on the storage volume ?

    Thanked by 2yongsiklee greentea
  • crunchbitscrunchbits Member, Patron Provider, Top Host
    edited February 2023

    @theyeti said:
    any chance of a restock of 4TB Storage VPS?

    Yes, it's coming. It will be posted here in this thread as soon as they're live.

    @andynz said:

    @crunchbits said:

    @andynz said:
    Hey guys, can you tell me which city the VPS's are physically in? My VPS IP resolves to Tampa Florida. Just interested in terms of distance/latency from me in NZ.

    Some of our newer IPv4 ranges show Tampa. They're not in Tampa, or anywhere near Florida. It's a geodb not-yet-updated thing it seems, if you check other sources it will show correctly (Cheney, WA or CDA, ID) but it's about 50/50.

    Great thanks. Explains similar latency to Seattle services I have.

    To labour a point :), just amazing service guys and great full featured control panel. Can the storage size be increased at a future date without losing current data on the storage volume ?

    I appreciate it. Regarding the storage size there isn't a nice/clean way to do it without wiping the existing. We have 2 ways that we're currently contemplating to make it work:

    1) Attach desired expanded storage size to customer's account, give them ~48hrs (or something reasonable) to migrate locally before we remove the smaller storage disk.
    2) Attach additional storage to add up to ideal overall storage required. I.E 2TB to 4TB plan results in /dev/vdb and /dev/vdc each of 2T.

    The unfortunate part at the moment is that we couldn't sell bigger storage anyways, but I assume at some point we'll outpace demand to at least get some breathing room.

  • Got customer feedback?

  • andynzandynz Member
    edited February 2023

    @crunchbits said:

    @theyeti said:
    any chance of a restock of 4TB Storage VPS?

    Yes, it's coming. It will be posted here in this thread as soon as they're live.

    @andynz said:

    @crunchbits said:

    @andynz said:
    Hey guys, can you tell me which city the VPS's are physically in? My VPS IP resolves to Tampa Florida. Just interested in terms of distance/latency from me in NZ.

    Some of our newer IPv4 ranges show Tampa. They're not in Tampa, or anywhere near Florida. It's a geodb not-yet-updated thing it seems, if you check other sources it will show correctly (Cheney, WA or CDA, ID) but it's about 50/50.

    Great thanks. Explains similar latency to Seattle services I have.

    To labour a point :), just amazing service guys and great full featured control panel. Can the storage size be increased at a future date without losing current data on the storage volume ?

    I appreciate it. Regarding the storage size there isn't a nice/clean way to do it without wiping the existing. We have 2 ways that we're currently contemplating to make it work:

    1) Attach desired expanded storage size to customer's account, give them ~48hrs (or something reasonable) to migrate locally before we remove the smaller storage disk.
    2) Attach additional storage to add up to ideal overall storage required. I.E 2TB to 4TB plan results in /dev/vdb and /dev/vdc each of 2T.

    The unfortunate part at the moment is that we couldn't sell bigger storage anyways, but I assume at some point we'll outpace demand to at least get some breathing room.

    Option 1 works for me and sounds the simplist. Are there any issues with resizing the primary 7GB to 15GB on the active system or is that a migrate too? I will watch out for when your catching up and upgrade fro mthe 2TB to 4TB then.

    OR

    I don't mind just signing up for a 4TB when they are avaiable and do the transfer myself then drop the m2m 2TB. It is easier and cleaner for you guys too.

    Thanked by 1greentea
  • @hkbbdx said:
    Got customer feedback?

    Right now, great UI (main website and control panel). Like that dark mode is supported (apart from brief flash before dark mode kicks in)

    2TB disk faster and slower than I expected. (As fast as the Online.net server I had, slower than the Servarica server I have).

    Control Panel significantly better than many I've had experience of. Obviously taken some time and care into what they provide users, and how it looks.

  • @andynz said:

    I don't mind just signing up for a 4TB when they are avaiable and do the transfer myself then drop the m2m 2TB. It is easier and cleaner for you guys too.

    I'll probably do the same. We don't need the extra storage, but I'm seeing lots of OOM activity trying to zfs recv 200+ GB datasets with only 1GB of RAM.

  • andynzandynz Member
    edited February 2023

    @pwned said:

    @andynz said:

    I don't mind just signing up for a 4TB when they are avaiable and do the transfer myself then drop the m2m 2TB. It is easier and cleaner for you guys too.

    I'll probably do the same. We don't need the extra storage, but I'm seeing lots of OOM activity trying to zfs recv 200+ GB datasets with only 1GB of RAM.

    Same here. The 1GB/1 core is having trouble at times with the migration I am doing. I wondered if it was node load but I am happy to upgrade to the 4TB with 2GB/2core when they are back in stock.

    Thanked by 1pwned
  • melp57melp57 Member
    edited February 2023

    Why is everything on your site out of stock(VPS) I was planning on getting a Ryzen VPS
    Guess I should have checked before adding funds to my account, which are non-refundable.

  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @tarasis said:

    @hkbbdx said:
    Got customer feedback?

    Right now, great UI (main website and control panel). Like that dark mode is supported (apart from brief flash before dark mode kicks in)

    2TB disk faster and slower than I expected. (As fast as the Online.net server I had, slower than the Servarica server I have).

    Control Panel significantly better than many I've had experience of. Obviously taken some time and care into what they provide users, and how it looks.

    Thank you for the feedback. Always appreciated.

    @pwned said:

    @andynz said:

    I don't mind just signing up for a 4TB when they are avaiable and do the transfer myself then drop the m2m 2TB. It is easier and cleaner for you guys too.

    I'll probably do the same. We don't need the extra storage, but I'm seeing lots of OOM activity trying to zfs recv 200+ GB datasets with only 1GB of RAM.

    @andynz said:

    @pwned said:

    @andynz said:

    I don't mind just signing up for a 4TB when they are avaiable and do the transfer myself then drop the m2m 2TB. It is easier and cleaner for you guys too.

    I'll probably do the same. We don't need the extra storage, but I'm seeing lots of OOM activity trying to zfs recv 200+ GB datasets with only 1GB of RAM.

    Same here. The 1GB/1 core is having trouble at times with the migration I am doing. I wondered if it was node load but I am happy to upgrade to the 4TB with 2GB/2core when they are back in stock.

    Is it just a memory thing, or are the vCPU amounts also limiting your use case? I think we might have a solution without forcing all of you into a paid upgrade or waiting period.

    @melp57 said:
    Why is everything on your site out of stock(VPS) I was planning on getting a Ryzen VPS
    Guess I should have checked before adding funds to my account, which are non-refundable.

    It’s out of stock because people bought them all and we’re not going to do preorders. Depending on the product you want, we always have some extra capacity live for hardware issues or emergencies. I’d also be happy to refund your credits.

    Thanked by 2tarasis pwned
  • @melp57 said:
    Why is everything on your site out of stock(VPS) I was planning on getting a Ryzen VPS
    Guess I should have checked before adding funds to my account, which are non-refundable.

    It’s out of stock because people bought them all and we’re not going to do preorders. Depending on the product you want, we always have some extra capacity live for hardware issues or emergencies. I’d also be happy to refund your credits.

    No need to refund, I can use for current server.

    Do you plan to restock Rezen?

  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @melp57 said:

    @melp57 said:
    Why is everything on your site out of stock(VPS) I was planning on getting a Ryzen VPS
    Guess I should have checked before adding funds to my account, which are non-refundable.

    It’s out of stock because people bought them all and we’re not going to do preorders. Depending on the product you want, we always have some extra capacity live for hardware issues or emergencies. I’d also be happy to refund your credits.

    No need to refund, I can use for current server.

    Do you plan to restock Rezen?

    Yes, Ryzens will have some limited availability again soon. Likely within 24 hours. You can also open a ticket from your existing account and I’d be happy to assign a Ryzen plan to your account

  • @crunchbits said:

    Is it just a memory thing, or are the vCPU amounts also limiting your use case? I think we might have a solution without forcing all of you into a paid upgrade or waiting period.

    So many OOM kills in the log that I honestly don't know if the CPU is also an issue. I'm running zfs elsewhere in even less ram just for fun, but in those locations the datasets are much, much smaller.

    Another factor might be that the initial zfs recv stream is huge due to hundreds of snapshots--I suspect 1GB would be fine once it's just the daily incremental recv. But then you have to wonder if 1GB would be enough to pull down the entire dataset in an emergency restore situation. That's a rare event (you'd probably just need the latest snapshot, not the entire dataset) so the workaround there would be to open a ticket begging for a temporary ram increase.

    Working through stuff like this is why I went monthly to start.

  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @pwned said:

    @crunchbits said:

    Is it just a memory thing, or are the vCPU amounts also limiting your use case? I think we might have a solution without forcing all of you into a paid upgrade or waiting period.

    So many OOM kills in the log that I honestly don't know if the CPU is also an issue. I'm running zfs elsewhere in even less ram just for fun, but in those locations the datasets are much, much smaller.

    Another factor might be that the initial zfs recv stream is huge due to hundreds of snapshots--I suspect 1GB would be fine once it's just the daily incremental recv. But then you have to wonder if 1GB would be enough to pull down the entire dataset in an emergency restore situation. That's a rare event (you'd probably just need the latest snapshot, not the entire dataset) so the workaround there would be to open a ticket begging for a temporary ram increase.

    Working through stuff like this is why I went monthly to start.

    Good info for us to consider on additional use cases. You can open a ticket and try now, see if it fixes your issues. No begging required 😂

    The hypervisors are over-provisioned with resources so short-term/initial setup quotas can be exceeded to help users migrate. Depending on how these are used and common issues, we could just look at doubling the RAM on each plan as well. Going forward not an issue, just requires short maintenance windows for everyone on the existing server.

    Thanked by 1maverick
  • @crunchbits said:

    @pwned said:

    @crunchbits said:

    Is it just a memory thing, or are the vCPU amounts also limiting your use case? I think we might have a solution without forcing all of you into a paid upgrade or waiting period.

    So many OOM kills in the log that I honestly don't know if the CPU is also an issue. I'm running zfs elsewhere in even less ram just for fun, but in those locations the datasets are much, much smaller.

    Another factor might be that the initial zfs recv stream is huge due to hundreds of snapshots--I suspect 1GB would be fine once it's just the daily incremental recv. But then you have to wonder if 1GB would be enough to pull down the entire dataset in an emergency restore situation. That's a rare event (you'd probably just need the latest snapshot, not the entire dataset) so the workaround there would be to open a ticket begging for a temporary ram increase.

    Working through stuff like this is why I went monthly to start.

    Good info for us to consider on additional use cases. You can open a ticket and try now, see if it fixes your issues. No begging required 😂

    The hypervisors are over-provisioned with resources so short-term/initial setup quotas can be exceeded to help users migrate. Depending on how these are used and common issues, we could just look at doubling the RAM on each plan as well. Going forward not an issue, just requires short maintenance windows for everyone on the existing server.

    My usage case is migrating a music collection from Hetzner to Crunchbit using rclone. I BW limited to 5MBs as a precaution and consideration to fellow node users. Do I need to do that? Also running is personal use Jellyfin media server which is scanning the new media. So overall I suspect my occasional overload is down to the 1GB/1Core on the 2TB plan. I did see system load of >5 at one point yesterday with a cpu load down around 40% and IO around 10% wait. I know it is not an equivalent. Basically waiting on the 4TB availabiltiy so I can move the 2Tb m2m over to that and drop the 2TB. I keep checking the store every few hours.

    Thanked by 2greentea maverick
  • Any storage servers stock soon? And is the promo code still active?

    Thanked by 1greentea
  • It’s out of stock

    >

    I had been putting off getting the storage deal for 2 reasons:

    1) hoping u wud introduce biannual and triennial deals :)
    2) double minded on the ram & the bandwidth. 1GB ram seems low imo.

    Should I keep my fingers crossed for a restock with double the ram and maybe double the bandwidth & triennial payment :smiley:

    Thanked by 1greentea
  • @andynz said:

    @crunchbits said:

    @pwned said:

    @crunchbits said:

    Is it just a memory thing, or are the vCPU amounts also limiting your use case? I think we might have a solution without forcing all of you into a paid upgrade or waiting period.

    So many OOM kills in the log that I honestly don't know if the CPU is also an issue. I'm running zfs elsewhere in even less ram just for fun, but in those locations the datasets are much, much smaller.

    Another factor might be that the initial zfs recv stream is huge due to hundreds of snapshots--I suspect 1GB would be fine once it's just the daily incremental recv. But then you have to wonder if 1GB would be enough to pull down the entire dataset in an emergency restore situation. That's a rare event (you'd probably just need the latest snapshot, not the entire dataset) so the workaround there would be to open a ticket begging for a temporary ram increase.

    Working through stuff like this is why I went monthly to start.

    Good info for us to consider on additional use cases. You can open a ticket and try now, see if it fixes your issues. No begging required 😂

    The hypervisors are over-provisioned with resources so short-term/initial setup quotas can be exceeded to help users migrate. Depending on how these are used and common issues, we could just look at doubling the RAM on each plan as well. Going forward not an issue, just requires short maintenance windows for everyone on the existing server.

    My usage case is migrating a music collection from Hetzner to Crunchbit using rclone. I BW limited to 5MBs as a precaution and consideration to fellow node users. Do I need to do that? Also running is personal use Jellyfin media server which is scanning the new media. So overall I suspect my occasional overload is down to the 1GB/1Core on the 2TB plan. I did see system load of >5 at one point yesterday with a cpu load down around 40% and IO around 10% wait. I know it is not an equivalent. Basically waiting on the 4TB availabiltiy so I can move the 2Tb m2m over to that and drop the 2TB. I keep checking the store every few hours.

    My high load ~1.3 appears to be purely down to WA time on the 2TB HDD. CPU is around 5-10% but WA is around 60% while the media files are being indexed. I'm guessing that's simply down to multi user accessing the storage backend.

  • @crunchbits said:

    @namein said:

    @crunchbits said: roughly around the $11 per TiB mark.

    the price is for per month? That was indeed a bit beyond expectations.

    I've had a great time with Crunchbits so far, and I'm impressed with the well-designed admin panel. Their technical support response is also very quick, and I think I'll be switching to an annual contract with them soon for the long haul.

    @yongsiklee said:

    ... We're exploring (based on some feedback and requests) an NVMe cached/accelerated HDD storage solution currently. We are also contemplating all-NVMe big storage plans--though they would be more expensive than HDD. I'm not sure what kind of appetite exists for those since they'll likely be roughly around the $11 per TiB mark.

    All NVMe storage? Do you see any demand for it in the near future? Otherwise, forget it unless you hit a jackpot.

    This wouldn't be a low-end focused offer. We do have users that require bigger storage (generally complimenting a more powerful VPS/Dedi) with speed requirements to match, but as far as I can tell SSDs are still plenty fast for their uses. The only difference is we actually get NVMe storage for less than SSDs now.

    This has been tied mostly to our GPU/HPC users. I guess sometimes(often?) their workloads outpace the ~140MB/s of spinning disk arrays.

    Yes! When working on trading systems, we frequently found that disk performance impacted our databases hard. Super fast, super reliable disks were crucial.

  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @andynz said:
    My usage case is migrating a music collection from Hetzner to Crunchbit using rclone. I BW limited to 5MBs as a precaution and consideration to fellow node users. Do I need to do that? Also running is personal use Jellyfin media server which is scanning the new media. So overall I suspect my occasional overload is down to the 1GB/1Core on the 2TB plan. I did see system load of >5 at one point yesterday with a cpu load down around 40% and IO around 10% wait. I know it is not an equivalent. Basically waiting on the 4TB availabiltiy so I can move the 2Tb m2m over to that and drop the 2TB. I keep checking the store every few hours.

    More good info. You don’t have to do that, but it is courteous. If any single users use too much (proportionally) of their disk for too long we'll alert and contact them privately. It hasn't been an issue yet. The systems have been relatively good at balancing, and I think more importantly at the price-point most customers seem to understand that it is a shared HDD array. We always sell NVMe and are known to be generous with bandwidth on request if it needs to be faster ;)

    @raza19 said:

    It’s out of stock

    >

    I had been putting off getting the storage deal for 2 reasons:

    1) hoping u wud introduce biannual and triennial deals :)
    2) double minded on the ram & the bandwidth. 1GB ram seems low imo.

    Should I keep my fingers crossed for a restock with double the ram and maybe double the bandwidth & triennial payment :smiley:

    Let's run the @raza19 special: You tell me the specs you want to see on the billing plan you want to see and if I can make it happen, we will run a special and dedicated some nodes to it. :)

    @bgerard said:
    Yes! When working on trading systems, we frequently found that disk performance impacted our databases hard. Super fast, super reliable disks were crucial.

    This is what has been relayed to us by commercial customers as well. It seems like common sense, but it was a lot more sensitive than even I thought.

  • any chance that the legendary storage (10x12) be available again? :p

    Thanked by 1greentea
  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @swat4 said:
    any chance that the legendary storage (10x12) be available again? :p

    Never again unless the ones who have them give them up or transfer them. We'll have some big dedi-storage stuff soon too, but not quite those E3's.

    Thanked by 1greentea
  • any plans on some EU location?
    preferably some based one like Norway, essentially as good as USA in terms of freedom of speech but much better with privacy.

    Thanked by 1greentea
  • >

    @raza19 said:

    It’s out of stock

    >

    I had been putting off getting the storage deal for 2 reasons:

    1) hoping u wud introduce biannual and triennial deals :)
    2) double minded on the ram & the bandwidth. 1GB ram seems low imo.

    Should I keep my fingers crossed for a restock with double the ram and maybe double the bandwidth & triennial payment :smiley:

    Let's run the @raza19 special: You tell me the specs you want to see on the billing plan you want to see and if I can make it happen, we will run a special and dedicated some nodes to it. :)

    >
    >

    Thank you for the honor :smiley:

    I wud grab this at lightening speed :joy:

    1 E5-2695 v4 vCore
    2GB DDR4 RAM
    7GB SSD OS/Primary Disk
    2TB HDD Storage Disk
    12TB Bandwidth @ 1Gbps
    1 IPv4 + 1 IPv6 /64
    Plus 15-20% additional discount for triennial payments 
    
    Thanked by 1greentea
  • Any ETA on 4TB storage units?

  • crunchbitscrunchbits Member, Patron Provider, Top Host

    @raza19 said:
    >

    @raza19 said:

    It’s out of stock

    >

    I had been putting off getting the storage deal for 2 reasons:

    1) hoping u wud introduce biannual and triennial deals :)
    2) double minded on the ram & the bandwidth. 1GB ram seems low imo.

    Should I keep my fingers crossed for a restock with double the ram and maybe double the bandwidth & triennial payment :smiley:

    Let's run the @raza19 special: You tell me the specs you want to see on the billing plan you want to see and if I can make it happen, we will run a special and dedicated some nodes to it. :)

    >
    >

    Thank you for the honor :smiley:

    I wud grab this at lightening speed :joy:

    > 1 E5-2695 v4 vCore
    > 2GB DDR4 RAM
    > 7GB SSD OS/Primary Disk
    > 2TB HDD Storage Disk
    > 12TB Bandwidth @ 1Gbps
    > 1 IPv4 + 1 IPv6 /64
    > Plus 15-20% additional discount for triennial payments 
    > 

    @andynz said:
    Any ETA on 4TB storage units?

    Not precise. There is a LET member or two helping us and giving feedback for certain use-cases so we can (freely) improve quality of services, if possible.

    One of those that is settled is a likely bump in RAM to all plans. I think the other is going to be adding some speed and IOPS headroom to the arrays. Basically, will be similar to what @raza19 put up (and even if it's just for @raza19 we'll add triennial payments ;) )

    Thanked by 3raza19 plumberg pwned
Sign In or Register to comment.