#How to prevent udp attack from OVH
137 messages · Page 1 of 1 (latest)
so what if i block all ovh's ip?
Then report the IPs for DDoS
the accounts involved will be suspended if intentional
can i block the netblock of ovh? I got it here
I'd only block abusive subnets
that list includes ovhcloud stuff
not only customer ranges
so its more risky
just block as they happen
you can check which IP belongs to which range
and block the range
remember to report the IPs too
this way it gets more expensive and hard for the attacker
he will have to use resellers and the resellers will blacklist him as they also get banned.
eventually he is left with no means to gain ovhcloud services to use with attacks.
the reports also allows abuse team to learn new type of attack traffic
this will then get fingerprinted and added to automatic suspension list eventually.
hmmm
So if I block both, what happens?
And where can I get ovh's ddos ips
at worst your routing goes bad or you lose WAN. I guess it won't happen though. Just disable monitoring to avoid restarts.
tcpdump the attack traffic as it happens with automatic script
then view IPs with wireshark
from the dump file.
I will try blocking both, but when the worst happens, will the OVH team support me for free? Or do I have to pay to get it fixed?
U there?
If I block both my OVH vps ip is it still good? or do i need to leave that netblock out?
Just try it
remove if it fails
You can't really permanently ruin anything
just reverse the changes if it fails.
yes
i only allow port 25565 (minecraft) and block all
but
I mean
does it work when the attack is coming from OVH?
"ovh's antiddos" will "not work" when "attacking from OVH"
I would assume that it includes all traffic, as those are public IPs too. You could also specifically add the offending subnets with ipv4 rule if that above fails.
but I can only add up to 20 rules
bruh
rule refuse ipv4 can it prevent udp attack from ovh?
so i'll try to block some of ovh's netbock
i know
setup automatic tcpdump
i try test
yes
This is what I use. https://github.com/Ikfes/AutomaticBashTCPDump
Its simple, and theres nothing extra
The code is less than 20 lines and easy to audit as long as you trust the packages it calls such as "tcpdump".
It simply records x amount of network traffic in packets when the attack starts and is on peak
and writes the record to .pcap file
whichever is the attacker, you can see with wireshark
it also shows the attack types if you know how to read it further.
20k?
you mean
20,000kb = 20mb?
Read the description in github
PPS
as in network packets
You wish to know the offending subnets, and that script is a tool to record any traffic during the attack to help you determine them.
Its not limited to OVH
You can use it with any linux and it can record any ip.
ok
but important
Whether IPv4 can prevent it from the OVH attack or not is important
Firewall network will it work if ddosed from OVH?
You can find all information in here 🙂
and you can always change the configurations and try different settings if something is not immediately working for you.
i tried and it's ok
Can I ask you something
Is there a way to let it remove the optional ips?
like i don't want it to capture ip 123.123.123.123
that is irrelevant as it only captures about 1 second or less during the attack
20k packets is pretty fast captured if theres an ongoing attack
you can edit the command though and exclude or include only some stuff or protocols if you check tcpdump command examples
but I see no use in that myself
generally you'd browse to middle of the file in wireshark and look whats the most recurring stuff in the file.
and ignore rest to determine whats the attack about
Ok
Hi @uneven agate
Refuse Protocol IPv4 is to block all protocol like TCP, UDP, ICMP, AH, GRE, ESP, ..... right?
Plz rep me
Correct
You can use that as priority 19
and set allow rules to priority 0-18
this way it only allows what you set to 0-18
and blocks everything else with 19 rule
thank u so much
No problem 🙂
Only 20 rules per IP is allowed for now
this may or may not change in future.
I do not know the reason myself, but if I had to guess, it has to do with performance. Your's IP is not the only one that has rules on it, and OVHCloud has millions of them with a lot of rules on many, and these rules are mirrored across the globe.
You can make best use of 20 rules fairly easy to achieve almost anything
You don't need to make many block rules in general. 1 is enough
Then have 19 rules for allowing whatever you need.
yes
i refuse all ipv4 priority 19
but the ddos ip will get in what i accept
ex: ip 123.123.123.123 attack port 25565 (i accepted port 25565)
and many other different ips
23.0000000
43.000000
...
How to report ip ovh ddos ?