[stunnel-users] Rapid subsequent DNS resolution fails within stunnel

[Vincent Deschenes]

Is it possible that if my http client keeps the connection alive with Stunnel then Stunnel would reuse variable/pointers which would have been cleared by previous call and therefore sends an empty name to the name server resulting in the EAI_NONAME error?

We are running with delay = yes option as the endpoint name change IP sometime.
We use Stunnel on the client side.

No one having that same problem ?


Thanks,

Vincent


From: stunnel-users [mailto:stunnel-users-bounces at stunnel.org] On Behalf Of Vincent Deschenes
Sent: Friday, 6 April 2018 11:35 AM
To: stunnel-users at stunnel.org
Subject: [stunnel-users] Rapid subsequent DNS resolution fails within stunnel


This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing<http://aka.ms/LearnAboutSpoofing>

Feedback<http://aka.ms/SafetyTipsFeedback>

Hi,

We have had this weird problem for a long time where sometime DNS resolution will fail with this error:

Error resolving 'domain name printed here': Neither nodename nor servname known (EAI_NONAME)

Today I found out that this will happen when stunnel get 2 calls for the same outgoing endpoint very fast one after the other.

One call will work no problem. But if you do 2 one after the other then the second one and all other subsequent will fail.
Wait a bit and try again, the first one will pass and then others will fail.

DNS resolution on the server (linux) itself work without problem, nslookup in loops no problem. But with stunnel we always have the problem.

The only working solution is to add a record to the server host file.

Any idea ? Could this be a bug in stunnel ?


Thanks,

Vincent


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stunnel.org/pipermail/stunnel-users/attachments/20180420/665fb268/attachment.html>