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.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
cool story
Yeah, they seem to be getting hammered over the last few days our wordpress blog hosted with wpengine in London is up and down like a yo yo.
That's called life.
They were have success because nobody threats them in past.
Shame, as for many providers here. What a shame, so big operator and died under ddos attacks. People will never learn basic things: security matters. The good side of the news about linode -> i'm sure, they will have ddos protection in 2016.
So Linode has been running all these years without any DDOS protection? That sounds pretty incredible considering they are a premium provider and many budget providers here have decent DDOS protections.
They used to have some form of nullrouting, i don't know whether it was automatic or manual. LET was hosted at Linode in the past and was getting nullrouted for 24 hours when DDoSed.
Incredible since they just bought a new HQ and paid $5 millions for it. They should've invested some of that money in better security.
Clients IP ranges being DDoSed like what was the reason to leave offers for most providers here.
The one time I was with linode I experienced 3 node failures in a month and they weren't quick solving the issues, always took them more than 5 hours.
If you consider a provider premium because they accurately inform you about their downtimes then you can find other providers who are a lot cheaper.
Refugee offers ? /s
Linode, KnownHost, WiredTree...none offer DDOS protection even as an addon.
In 2015 (soon to be 2016) they should have the infrastructure in place to handle these attacks --- if they dont, it is time to move on.
None of them sound exceptionally quality to justify their prices.
Thanks
Price of DDoS equipment has dropped a lot
haha
I've used all three. Linode was meh though their panel had some unique features in its day.
WT support was AWESOME. And out of the box, their servers are very secure - they really go the extra mile to provision cPanel with lots of security enhancements, custom mod_security rules, etc. I really liked WT and if I was a semi-technical user who wanted hand-holding support and easy-to-use managed cPanel, I would use them.
KH support was almost as good and out of the box, they put in some good stuff, though not like WT. With KH you get more RAM per plan, while with WT you get more disk.
But none of them have DDOS protection, which just boggles my mind. They could be offering it as a premium add-on.
They got popular by not being total garbage and advertising heavily on WHT getting recommended by users for months.
Not that you are generally wrong, but I received very good support and service from both WiredTree and KnownHost in the past. They were the right providers at the right time for me. Now, that I can manage servers myself, they seem superfluous, but looking back, both did a good job.
I was hesitating to open a thread about this hehe...
For the capacity of the links Linode claims to have, these should be really big attacks. Unfortunately the servers at my work use the Dallas region, which has been the most affected (until today, which seems to be London).
This time doesn't seem like a client getting ddos'd, but some haters attacking the whole linode infraestructure.
I was a little hesitant at first but seems as the attacks are so heavy I thought never mind. They will surely have to bring in DDOS protection now. If not I'll have to move servers.
I was considering to move some important projects, at least to another datacenter, but unfortunately you can't carry your IP addresses across DCs as far as I know. This is important for my particular use cases =(
Price of anti-ddos hardware is 80% software and 20% ASICs, SW pricing is the same as usual but 10/40G ASICs got cheaper last year(s).
Yes, only null and enough upstream capacity.
In some linode locations 100Gbit BW - which is already low on protection these days - costs a million PER MONTH, then you also need the HW on top to actually use that...
Well, that's interesting how much OVH invested to their anti-ddos protection.
Arbor + Tilera + custom in-house : https://www.ovh.com/us/anti-ddos/
Passes through a few filter layers, plus all of the BW they have to absorb around their PoPs.
https://www.ovh.com/us/news/articles/a1171.protection-anti-ddos-service-standard says it costs them around $10 million using traditional filtering, and $103,000 for 100gbit using their own technology.
If only Linode invested that $5 million they spent buying a building for only 160 people just because the CEO liked it back into their own platform...
BW in France/CA is 10x cheaper than in JP.
This figure does not include any BW.
I build you 40G filtering at 25k as well, and i need 5k alone for the 40G network cards.
Which operator/site keep online after targeted by ddoser?
Come on, you pay ~ $10-$20/m including ddos protection, if you compare to ovh, they will beat all provider in pricing + ddos protection. Why none complaint about delimiter, qps, datashack ddos protection?
http://status.linode.com/incidents/mmdbljlglnfd
"Additionally, we will be announcing the details of an ongoing project to significantly improve our internet connectivity and resiliency. "
I don't see them rolling out filtering, I mean, they don't even have their own routers in each of their locations. If you lookup their IP space you'll see that it's being announced by their upstream datacenters and or they peer over a private ASN. If they're having this much of an issue dealing with floods against their platform now, do you think they'll have the time & manpower to deal with leaks on filtering that will happen?
When LE* was getting hammered back in the day the floods were around ~500mbit. Linode still nullrouted over it, though. The attacker came after us at one point but it never got over a gbit sustained.
I'm not sure why they rolled out 40gbit uplinks on all of their servers when they obviously can't support it at the core. Sure, you can get some nice (< 2gbit) download speeds on the VM's, but it sounds like they only have a couple 10gigs at the core.
We'll see, maybe I'm completely wrong and they're eating some cloudflare sized attacks, but I don't see softlayer/coloat sitting on 300~400gbit of capacity in those locations to get knocked offline.
Francisco
After 189 days, my Newark node kept going on, and NAC is brilliant (8001) - I've colo'd there myself, and they have plenty of bandwidth.
I honestly don't know what's going on, it's targeted as you saw on WHT Dallas was down right before Newark went down. Some pissed off former employee or pathetic local competitor would be my guess.
Had to move things to OVH after 1/2 year perfect. From dealing with NAC myself over the years with colo space & transit, 8001 is where you want to be for unparalleled transit, peering, and stability for the east coast. NAC started as a dial up in the 90's, so they have multiple PoPs, LINX peering, etc.
Not sure where the ball is dropped between robust NAC network -> Linode, but they don't have their own ASN, using 8001, and the prefixes were gone from the route-tables.
That's when I decided to move. Not even bothering trying to mitigate in RT, but just dropped from the net, all subnets were null-routed.