5 ms·

Always consider if 12 hours of lost revenue is worth the savings. Recently hetzner has been flakey with minimum or no response for support or even status update

Always consider if 12 hours of lost revenue is worth the savings. Recently hetzner has been flakey with minimum or no response for support or even status updates that anything was wrong. My favorite was them blaming an issue on my side just to have a maintenance status update the day after about congestion.

arbol 1y ago
Atlas wasn't giving us any support for $3K per month. Hetzner at least have some channel to contact them, which is an improvement. That said, if their uptime is rubbish them we'll probably migrate again. Moving back to Atlas is not an option as we were getting hammered by the data transfer costs and this was only going to increase due to our architecture. Thanks for reading!
- zamalek 1y ago
  OVH is allegedly pretty good. I host all my personal stuff on Hetzner right now so I can't speak to it personally.
  - arbol 1y ago
    We also use OVH and have so far not had any downtime in about 6 months.
- kosherhurricane 1y ago
  500GB isn't a lot of data, and $3K/month seems like an extortion for that little data. Having said that, MongoDB pricing page promises 99.995% uptime, which is outstanding, and would probably be hard to beat that doing it oneself, even after adding redundancy. But maybe you don't need that much uptime for your particular use case.
  - arbol 1y ago
    Its more like 700GB now on the new server and we were about to have to migrate to a higher tier on Atlas. > maybe you don't need that much uptime for your particular use case. Correct. Thanks for reading!
    - tecleandor 1y ago
      Yep, we just migrated to Atlas, and the disk size limitation of the lower instance tiers pushed us to do a round of data cleaning before the migration. Also, we noticed that after migration, the databases that were occupying ~600GB of disk in our (very old) on premise deployment, were around 1TB big on Atlas. After talking with support for a while we found that they were using Snappy compression with a relatively low compression level and we couldn't change that by ourselves. After requesting it through support, we changed to zstd compression, rebuilt all the storage, and a day or two later our storage was under 500GB. And backup pricing is super opaque. It doesn't show concrete pricing on the docs, just ranges. And depending on the cloud you deployed, snapshots are priced differently so you can't just multiply you storage by the number of the snapshots, and they aren't transparent about the real size of the snapshots. All the storage stuff is messy and expensive...
    - gervwyk 1y ago
      Did you have a very aggressive backup schedule?
  - buster 1y ago
    That's all fine and such, but i suppose the SLAs aren't covering your revenue loss. In fact after looking at https://www.mongodb.com/legal/sla/atlas/data-federation#:~:text=You%20must,support%20team. https://www.mongodb.com/legal/sla/atlas/data-federation#:~:t... it makes me wonder how much worth the SLA is. 10% Service Credit after all the limitations? Atlas can take their 10% Service Credit, i wouldn't care. Save the money and chose a stable provider.
  - gizzlon 1y ago
    > Having said that, MongoDB pricing page promises 99.995% uptime Or.. what? That's the important part
0x073 1y ago
Using hetzner since 5 years never had issues and only 1 downtime in one data center.
- kachapopopow 1y ago
  I think the issue stems from their poor cloud infrastructure since that's where I've had the most issues the dedicated servers seem fine, that being said 2 years prior I had no issues either so it's definitely something recent.
  - 0x073 1y ago
    Which region? For me: 2 dedicated, 3 vps (nbg1) 1 vps (fsn1) Fsn is relatively new.
izacus 1y ago
My Hetzner instances all have higher reliability and uptime than AWS deployments. For years now. That was an interesting surprise.
- jabwd 1y ago
  Curious what kind of deployments you are running with them? I only have personal stuff with Hetzner; but never had issues so far (bare metal in my case coz cheap for what I get and need).
  - izacus 11mo ago
    Mostly EC2 type VMs with docker clients in them. Keeping infrastructure simple is important for us :)
mbesto 1y ago
If I understand correctly, the author's company provides a CAPTCHA alternative, which presumably means that if their service goes down, all of their customer's logins, forms, etc. either become inoperable or don't provide the security the company is promising by using their service. This makes me want to use the company's service less because now I know they can't survive an outage in a consistent and resilient way.
- goastler 11mo ago
  We have 14 servers and counting running the frontend captcha service, we have a high uptime which you can observe here: https://portal.prosopo.io/status https://portal.prosopo.io/status We have extra provisions for enterprise clients to provide rock solid SLAs for every use case. The db in question is our data store for event that we use for aggregated features such as traffic analysis and ml. This service lags behind our realtime services so we can deal with some downtime if necessary

arbol 1y ago

Atlas wasn't giving us any support for $3K per month. Hetzner at least have some channel to contact them, which is an improvement. That said, if their uptime is rubbish them we'll probably migrate again. Moving back to Atlas is not an option as we were getting hammered by the data transfer costs and this was only going to increase due to our architecture. Thanks for reading!

zamalek 1y ago
OVH is allegedly pretty good. I host all my personal stuff on Hetzner right now so I can't speak to it personally.
- arbol 1y ago
  We also use OVH and have so far not had any downtime in about 6 months.
kosherhurricane 1y ago
500GB isn't a lot of data, and $3K/month seems like an extortion for that little data. Having said that, MongoDB pricing page promises 99.995% uptime, which is outstanding, and would probably be hard to beat that doing it oneself, even after adding redundancy. But maybe you don't need that much uptime for your particular use case.
- arbol 1y ago
  Its more like 700GB now on the new server and we were about to have to migrate to a higher tier on Atlas. > maybe you don't need that much uptime for your particular use case. Correct. Thanks for reading!
  - tecleandor 1y ago
    Yep, we just migrated to Atlas, and the disk size limitation of the lower instance tiers pushed us to do a round of data cleaning before the migration. Also, we noticed that after migration, the databases that were occupying ~600GB of disk in our (very old) on premise deployment, were around 1TB big on Atlas. After talking with support for a while we found that they were using Snappy compression with a relatively low compression level and we couldn't change that by ourselves. After requesting it through support, we changed to zstd compression, rebuilt all the storage, and a day or two later our storage was under 500GB. And backup pricing is super opaque. It doesn't show concrete pricing on the docs, just ranges. And depending on the cloud you deployed, snapshots are priced differently so you can't just multiply you storage by the number of the snapshots, and they aren't transparent about the real size of the snapshots. All the storage stuff is messy and expensive...
  - gervwyk 1y ago
    Did you have a very aggressive backup schedule?
- buster 1y ago
  That's all fine and such, but i suppose the SLAs aren't covering your revenue loss. In fact after looking at https://www.mongodb.com/legal/sla/atlas/data-federation#:~:text=You%20must,support%20team. https://www.mongodb.com/legal/sla/atlas/data-federation#:~:t... it makes me wonder how much worth the SLA is. 10% Service Credit after all the limitations? Atlas can take their 10% Service Credit, i wouldn't care. Save the money and chose a stable provider.
- gizzlon 1y ago
  > Having said that, MongoDB pricing page promises 99.995% uptime Or.. what? That's the important part

0x073 1y ago

Using hetzner since 5 years never had issues and only 1 downtime in one data center.

kachapopopow 1y ago
I think the issue stems from their poor cloud infrastructure since that's where I've had the most issues the dedicated servers seem fine, that being said 2 years prior I had no issues either so it's definitely something recent.
- 0x073 1y ago
  Which region? For me: 2 dedicated, 3 vps (nbg1) 1 vps (fsn1) Fsn is relatively new.

izacus 1y ago

My Hetzner instances all have higher reliability and uptime than AWS deployments. For years now. That was an interesting surprise.

jabwd 1y ago
Curious what kind of deployments you are running with them? I only have personal stuff with Hetzner; but never had issues so far (bare metal in my case coz cheap for what I get and need).
- izacus 11mo ago
  Mostly EC2 type VMs with docker clients in them. Keeping infrastructure simple is important for us :)

mbesto 1y ago

If I understand correctly, the author's company provides a CAPTCHA alternative, which presumably means that if their service goes down, all of their customer's logins, forms, etc. either become inoperable or don't provide the security the company is promising by using their service. This makes me want to use the company's service less because now I know they can't survive an outage in a consistent and resilient way.

goastler 11mo ago
We have 14 servers and counting running the frontend captcha service, we have a high uptime which you can observe here: https://portal.prosopo.io/status https://portal.prosopo.io/status We have extra provisions for enterprise clients to provide rock solid SLAs for every use case. The db in question is our data store for event that we use for aggregated features such as traffic analysis and ml. This service lags behind our realtime services so we can deal with some downtime if necessary