Cases like that (even if the CEO takes pity and resolves it) are why I will always prefer to host on a VPS with a hard egress limit. Hit the limit (legitimately or from DDoS), network access is suspended until you take action (either wait for next billing cycle or buy another 10 TB for like $5-10). Can still log into virtual console to investigate, setup mitigations, etc.
Or better, unmetered. OVH might be a bit of a mess in many areas, but my server is unmetered. Doesn’t matter if a VM starts mining crypto or if I get DDoS’d or someome just wants to waste my bandwidth. Network can be pegged 24/7 for all I care, same price in the end.
Hosting companies know they can make a lot of money with on demand pricing like that, and they love it because for the most part you can’t do anything about it. If this was a company and not an individual, and the CEO didn’t have pity, I’m sure they’d have tried their best to extract that 5k, maybe even 20k or whatever the sales representative thinks they can get out of you. It’s crazy how the discounts become plentiful when it’s obvious there’s no way you can pay it all.
I’ve heard mixed reviews about OVH and heard they over-provision really badly. Does that cause you any issues / is that inaccurate in your experience?
Only one of my VPS’s is truly unmetered, and it’s the one I’ve had since like 2013 and is very, very grandfathered into a lot of perks . lol. I’m holding on to that one as long as I can. It’s also got a squeaky-clean IPv4 address that has only ever been used by me.
The rest of the ones I run have a fairly high cap and only meter egress traffic. I think they’re like 4 or 5 TB/mo on most of my plans. I’ve never hit anywhere near that limit even with one of them acting as my Lemmy CDN. Highest I’ve ever gotten was 75% which was enough to trigger a warning email.
I have a dedicated server, so they can’t possibly overprovision that. Just load up the OS over IPMI and I manage the VMs and all. Been using them before AWS was a thing and couldn’t be bothered, I like having all the control I can. I have a nice /29 of clean IPs with it that I’ve owned for 8 years as well.
OVH’s IPv6 is total ass though. Don’t even try, it’s essentially unusable especially if you want to use more than like 8 single IPs of your /56. The routers crap themselves and forget about the rest because it’s not a routed prefix, it sees it as if you have a single box claiming 8 IPs on itself.
I’m not sure I would use their cloud offerings. Renting old baremetal from them for cheap is much more price effective, especially if you can snag it on sale. And it also reduces waste by stopping those old boxes from being trashed and putting them to good use.
I suggest checking out their discount brands Kimsufi and SoYouStart. I pay like C$12/month for a dedicated server with a few cores, 8GB of RAM, and 2TB of hard drive space.
They used to link to my dig wrapper on my homepage for having their clients debug DNS problems for many years - even with translations of my UI in the various language help sites. I always found it amusing that a hoster of their size does that, instead of spending a lunchbreak to throw something together that integrates with their help page.
There also was a non significant number of users which didn’t understand that my homepage had nothing to do with OVH, and ended up mailing me about their DNS problems.
Yeah but then you have a customer calling and screaming at you “We just launched our big sale of the year and our site has been down for an hour!!!”.
If you let them burst and bill them, you end up with angry clients. If you don’t, you end up with angry clients. Letting them burst and being forgiving with the bill is the better approach IMHO.
You advice probably doesn’t apply to the OP in the image, as a “simple static site” is probably their blog or project wiki. It’s very unlikely they even have clients. For that case just having a hard limit and waiting is much safer.
I mean, I get email notifications as I’m approaching the threshold so I’m never caught off guard unless I ignore those. If everything’s legit (e.g. no DDoS), I can just add extra egress bandwidth with no interruptions.
Almost like it’d be trivial to do both, have it configured as a user setting, or even an option to say “charge up to X before contacting me, and Y before shutting down”
Cases like that (even if the CEO takes pity and resolves it) are why I will always prefer to host on a VPS with a hard egress limit. Hit the limit (legitimately or from DDoS), network access is suspended until you take action (either wait for next billing cycle or buy another 10 TB for like $5-10). Can still log into virtual console to investigate, setup mitigations, etc.
Or better, unmetered. OVH might be a bit of a mess in many areas, but my server is unmetered. Doesn’t matter if a VM starts mining crypto or if I get DDoS’d or someome just wants to waste my bandwidth. Network can be pegged 24/7 for all I care, same price in the end.
Hosting companies know they can make a lot of money with on demand pricing like that, and they love it because for the most part you can’t do anything about it. If this was a company and not an individual, and the CEO didn’t have pity, I’m sure they’d have tried their best to extract that 5k, maybe even 20k or whatever the sales representative thinks they can get out of you. It’s crazy how the discounts become plentiful when it’s obvious there’s no way you can pay it all.
I’ve heard mixed reviews about OVH and heard they over-provision really badly. Does that cause you any issues / is that inaccurate in your experience?
Only one of my VPS’s is truly unmetered, and it’s the one I’ve had since like 2013 and is very, very grandfathered into a lot of perks . lol. I’m holding on to that one as long as I can. It’s also got a squeaky-clean IPv4 address that has only ever been used by me.
The rest of the ones I run have a fairly high cap and only meter egress traffic. I think they’re like 4 or 5 TB/mo on most of my plans. I’ve never hit anywhere near that limit even with one of them acting as my Lemmy CDN. Highest I’ve ever gotten was 75% which was enough to trigger a warning email.
I have a dedicated server, so they can’t possibly overprovision that. Just load up the OS over IPMI and I manage the VMs and all. Been using them before AWS was a thing and couldn’t be bothered, I like having all the control I can. I have a nice /29 of clean IPs with it that I’ve owned for 8 years as well.
OVH’s IPv6 is total ass though. Don’t even try, it’s essentially unusable especially if you want to use more than like 8 single IPs of your /56. The routers crap themselves and forget about the rest because it’s not a routed prefix, it sees it as if you have a single box claiming 8 IPs on itself.
I’m not sure I would use their cloud offerings. Renting old baremetal from them for cheap is much more price effective, especially if you can snag it on sale. And it also reduces waste by stopping those old boxes from being trashed and putting them to good use.
Going to have to look into that. I didn’t know that was even an option. Good tip!
In all fairness, they usually end up on eBay and then in my basement 😂
I suggest checking out their discount brands Kimsufi and SoYouStart. I pay like C$12/month for a dedicated server with a few cores, 8GB of RAM, and 2TB of hard drive space.
They used to link to my dig wrapper on my homepage for having their clients debug DNS problems for many years - even with translations of my UI in the various language help sites. I always found it amusing that a hoster of their size does that, instead of spending a lunchbreak to throw something together that integrates with their help page.
There also was a non significant number of users which didn’t understand that my homepage had nothing to do with OVH, and ended up mailing me about their DNS problems.
Yeah but then you have a customer calling and screaming at you “We just launched our big sale of the year and our site has been down for an hour!!!”.
If you let them burst and bill them, you end up with angry clients. If you don’t, you end up with angry clients. Letting them burst and being forgiving with the bill is the better approach IMHO.
You advice probably doesn’t apply to the OP in the image, as a “simple static site” is probably their blog or project wiki. It’s very unlikely they even have clients. For that case just having a hard limit and waiting is much safer.
Is there a customer involved here?
After all if it’s for a customer it might be better to just give them the choice since the bill is on them.
I mean, I get email notifications as I’m approaching the threshold so I’m never caught off guard unless I ignore those. If everything’s legit (e.g. no DDoS), I can just add extra egress bandwidth with no interruptions.
Almost like it’d be trivial to do both, have it configured as a user setting, or even an option to say “charge up to X before contacting me, and Y before shutting down”