Don't receive http requests from some Ukrainian providers?

Question

Computer networks

An interesting situation has arisen. From September 21, we started receiving complaints about the inaccessibility of our sites from some Ukrainian providers, most often it was "VOLIA". Our services are located in the RTComm data center. Traces and pings pass normally, but http\https requests do not reach. Removing tcpdump shows the following:
tcpdump: listening on eth1
13:37:10.177110 client.52706 > server.80: S [tcp sum ok] 2115827637:2115827637(0) win 8192 <mss 1460,nop,wscale 8,nop,nop, sackOK> (DF) [tos 0x70] (ttl 119, id 30736, len 52)
13:37:10.177321 server.80> client.52706: S [tcp sum ok] 3382711418:3382711418(0) ack 2115827638 win 8192 <mss 1400,nop,wscale 8,nop,nop,sackOK> (DF) (ttl 127, id 22741, len 52)
13:37:10.260386 client.52706 > server.80:. [tcp sum ok] 1:1(0) ack 1 win 257 (DF) [tos 0x70] (ttl 119, id 30746, len 40)
13:37:29.154873 client.52706 > server.80: R [tcp sum ok ] 276:276(0) ack 1 win 0 (DF) [tos 0x70] (ttl 119, id 32459, len 40)
13:37:29.155106 server.80 > client.52706:. [tcp sum ok] 1:1(0) ack 1 win 257 (DF) (ttl 127, id 29721, len 40)
This shows that when the break occurs, the client transmits with an offset of 276, somewhere the size of the http request is obtained, but up to us he didn't get there. What could be?

Answer 1

I don’t know what exactly happened to you,
BUT
WILL, this is a hell of a pipets ,
for three weeks I twirled the brains of their those supporters
until it dawned on them that they had the same cache in the DNS for the third week