AWS Lightsail has good uptime for me. Inception as well (they run the forum here, right?). Currently trying Netcup and it’s been good so far. Hetzner Cloud is also pretty good from what I saw. Linode seems good too. Managing a client’s server there and didn’t see an issue in 2 years.
Most providers that have been around for a while have good uptime, I’d say.
My experience, a lot of providers' performance varies by location. High uptime in no particular order, Inception Amsterdam, UltraVPS.eu Düsseldorf, Avoro Frankfurt, and (although they get bagged a lot) some VirMach locations are objectively very reliable.
My best two since implementing my current monitoring solution (Freshping) in October 2021 are Primcast and Oracle Cloud Free Tier. Both have either had outages lasting less than 5 minutes (too fast for my monitoring to catch) or have not had any outages at all.
Some hosts I've had downtime with, from best to worst since implementing in October 2021 (provider names have been censored to not hurt anyone's feelings) are as follows:
Note that these don't specify reason; they could be due to a network, host issue, or admin error
My biggest surprise was the I*** VPS I have. I didn't expect it to have so much downtime, but I honestly can't complain as it was a Black Friday special, and I don't need it 100% of the time. To be fully honest, I have the same mentality with all my stuff. It's all so widely distributed across providers, datacenters, and geographical locations that a single outage doesn't affect me that much. In terms of feel I'd definitely say I've accomplished 99.95% uptime from my own perspective; everything has always been available when I wanted/needed it, so I'm pretty happy overall. Just surprised by some of the metrics!
@CamoYoshi said:
My biggest surprise was the I*** VPS I have. I didn't expect it to have so much downtime
I don't mind naming names. My Inception VPS in Phoenix is worse than their other locations (London, Amsterdam), which I suspect is because they use a different (non-Clouvider) provider in that location. Similarly, UltraVPS.eu in Los Angeles uses Psychz and this has not been as good (in my experience) as their Germany stuff.
The other thing is that when we talk about 99.9x% availability a single event can skew the statistics unless they cover several years. For example, your GreenCloud (I assume) was 99.8%. I've got 4 services with them and (without checking) 3 have had no downtime whereas the other had some hardware failure and is 99.8%. So averaged across all 4 the number is 99.95%, but if I only had one VPS my impression might be good or bad depending on which one it is. The higher the expected reliability, the longer we really need to measure to get meaningful numbers that average out random events.
So knowing who is reliable and sharing thoughts with each other is good, but also nice to consider the big picture - crazy stuff can happen to any provider, and the question is how do they deal with it. I've also said before that I'd rather a provider that has a solid 24 hours of downtime over one which has an unpredictable 24 x 1 hour, because I can mitigate that much more easily.
@CamoYoshi said:
My biggest surprise was the I*** VPS I have. I didn't expect it to have so much downtime
I don't mind naming names.
Just didn't want to come across as overly critical but you are right, I think your mentality is a good one to have. Being honest is important. Just was concerned about the inferred reasoning of my post rather than just providing cold hard data. Great points!
Cheap dedis are my drug, and I'm too far gone to turn back.
@CamoYoshi said:
My biggest surprise was the I*** VPS I have. I didn't expect it to have so much downtime
I don't mind naming names.
Just didn't want to come across as overly critical but you are right, I think your mentality is a good one to have. Being honest is important. Just was concerned about the inferred reasoning of my post rather than just providing cold hard data. Great points!
Perfectly understandable! Raw data is always good.
WordPress.com - hands down N01 in terms of uptime and stability.
Not my choice for the lack of freedom and prices, but hard to beat in terms of uptime.
HostMantis comes second, of those I've used and tested.
No idea whether their VPS hosting is any good though, only used shared and reseller.
@vitobotta said: How do handle kernel and package upgrades that require reboots? do you keep your servers up to date?
That one, obviously, isn't as up to date as it should be. But it also doesn't do anything production (although that's still no excuse). I've been toying with cancelling it for a while but part of me wanted to see how high it went lol.
Comments
We use Hetrix
PeekaBoo at https://handyhost.net
(Not affiliated with the Russian Handyhost Website in any way.)
Hetrix?
hetrixtools.com
PeekaBoo at https://handyhost.net
(Not affiliated with the Russian Handyhost Website in any way.)
It's an uptime monitor. I'm looking for good uptime proviers.
Hostolutions
VPS reviews and benchmarks |
Oh my bad misunderstood your post
PeekaBoo at https://handyhost.net
(Not affiliated with the Russian Handyhost Website in any way.)
kidding me?
Hostolutions.ro? isn't is deadpool?
You did not define parameters for uptime.
My criterion for best uptime is different, where host solutions triumphs.
VPS reviews and benchmarks |
take it easy, just run
uptime
command is enough for me$ uptime
@uptime
nice lol
AWS Lightsail has good uptime for me. Inception as well (they run the forum here, right?). Currently trying Netcup and it’s been good so far. Hetzner Cloud is also pretty good from what I saw. Linode seems good too. Managing a client’s server there and didn’t see an issue in 2 years.
Most providers that have been around for a while have good uptime, I’d say.
Simple and secure Borg Backup hosting from $2/month: BorgBase.com
Bytemark (now owned by iomart) have been pretty reliable. Odd network blip, but...
My experience, a lot of providers' performance varies by location. High uptime in no particular order, Inception Amsterdam, UltraVPS.eu Düsseldorf, Avoro Frankfurt, and (although they get bagged a lot) some VirMach locations are objectively very reliable.
Plus Azure & the like.
So far, Darkstar has 100% uptime between reboots for kernel updates, of which there have been several in recent days. The most recent was:
Right now:
¡Saludos! 🌎🌍
MetalVPS
How do handle kernel and package upgrades that require reboots? do you keep your servers up to date?
My best two since implementing my current monitoring solution (Freshping) in October 2021 are Primcast and Oracle Cloud Free Tier. Both have either had outages lasting less than 5 minutes (too fast for my monitoring to catch) or have not had any outages at all.
Some hosts I've had downtime with, from best to worst since implementing in October 2021 (provider names have been censored to not hurt anyone's feelings) are as follows:
W*** - Hillsboro, Oregon, USA location:
1hr, 5m - Avg Uptime 99.97%
O*** - Amsterdam, Netherlands location:
3hr, 36m - Avg Uptime 99.91%
G*** - Osaka, Japan location:
8hr, 6m - Avg Uptime 99.80%
I*** - Phoenix, Arizona, USA location:
12hr, 21m - Avg Uptime 99.72%
M*** - Vienna, Austria location:
22hr, 23m - Avg Uptime 99.36%
Note that these don't specify reason; they could be due to a network, host issue, or admin error
My biggest surprise was the I*** VPS I have. I didn't expect it to have so much downtime, but I honestly can't complain as it was a Black Friday special, and I don't need it 100% of the time. To be fully honest, I have the same mentality with all my stuff. It's all so widely distributed across providers, datacenters, and geographical locations that a single outage doesn't affect me that much. In terms of feel I'd definitely say I've accomplished 99.95% uptime from my own perspective; everything has always been available when I wanted/needed it, so I'm pretty happy overall. Just surprised by some of the metrics!
Cheap dedis are my drug, and I'm too far gone to turn back.
I don't mind naming names. My Inception VPS in Phoenix is worse than their other locations (London, Amsterdam), which I suspect is because they use a different (non-Clouvider) provider in that location. Similarly, UltraVPS.eu in Los Angeles uses Psychz and this has not been as good (in my experience) as their Germany stuff.
The other thing is that when we talk about 99.9x% availability a single event can skew the statistics unless they cover several years. For example, your GreenCloud (I assume) was 99.8%. I've got 4 services with them and (without checking) 3 have had no downtime whereas the other had some hardware failure and is 99.8%. So averaged across all 4 the number is 99.95%, but if I only had one VPS my impression might be good or bad depending on which one it is. The higher the expected reliability, the longer we really need to measure to get meaningful numbers that average out random events.
So knowing who is reliable and sharing thoughts with each other is good, but also nice to consider the big picture - crazy stuff can happen to any provider, and the question is how do they deal with it. I've also said before that I'd rather a provider that has a solid 24 hours of downtime over one which has an unpredictable 24 x 1 hour, because I can mitigate that much more easily.
Just didn't want to come across as overly critical but you are right, I think your mentality is a good one to have. Being honest is important. Just was concerned about the inferred reasoning of my post rather than just providing cold hard data. Great points!
Cheap dedis are my drug, and I'm too far gone to turn back.
VPSSLIM and LiteServer have been solid for me
Perfectly understandable! Raw data is always good.
Vultr and DO. I have some instances with them with 1500+ days uptime. I also have a good experience with Aruba at the past, with 1250+ days uptime.
WordPress.com - hands down N01 in terms of uptime and stability.
Not my choice for the lack of freedom and prices, but hard to beat in terms of uptime.
HostMantis comes second, of those I've used and tested.
No idea whether their VPS hosting is any good though, only used shared and reseller.
Detailed info about providers whose services I've used:
BikeGremlin web-hosting reviews
That one, obviously, isn't as up to date as it should be. But it also doesn't do anything production (although that's still no excuse). I've been toying with cancelling it for a while but part of me wanted to see how high it went lol.
terrahost. Holy quaga molly do they are stable! My VPN pumping ISO's 24/7 and is under heavy usage for *tube browsing and still going strong.
Uptime as in 'uptime' command: AWS (>3y), Azure (~1y)
Uptime as in external connectivity measured with Hetrix 1m ping intervals:
Contribute your idling VPS/dedi (link), Android (link) or iOS (link) devices to medical research
With over a decade of use, Ramnode is the most stable provider we've ever seen. Second would probably be Tilaa.