Sporadic Timeouts after upgrading to bind9.20

[Ondřej Surý]

Hi Klaus,

we've identified an issue in the glue cache that have been causing drops in the performance.

Can you test a development branch or do you need fix on top of 9.20?

Ondrej
--
Ondřej Surý (He/Him)
ondrej at isc.org

My working hours and your working hours may be different. Please do not feel obligated to reply outside your normal working hours.

> On 9. 9. 2024, at 10:39, Klaus Darilion via bind-users <bind-users at lists.isc.org> wrote:
> 
> As we still have several timeouts I downgraded our server to 9.18. If you know another workaround or need someone to test new version please let me know.
>  
> Thanks
> Klaus
>  
> From: Klaus Darilion <klaus.darilion at nic.at>
> Sent: Saturday, September 7, 2024 12:36 AM
> To: Klaus Darilion <klaus.darilion at nic.at>; Ondřej Surý <ondrej at isc.org>
> Cc: Klaus Darilion via bind-users <bind-users at lists.isc.org>
> Subject: RE: Sporadic Timeouts after upgrading to bind9.20
>  
> Correcting myself: event with { reuseport no; };  and UV_THREADPOOL_SIZE=12 still timeouts happen, but the situation improved a lot.
> Regards
> Klaus
>  
> From: bind-users <bind-users-bounces at lists.isc.org <mailto:bind-users-bounces at lists.isc.org>> On Behalf Of Klaus Darilion via bind-users
> Sent: Saturday, September 7, 2024 12:21 AM
> To: Ondřej Surý <ondrej at isc.org <mailto:ondrej at isc.org>>
> Cc: Klaus Darilion via bind-users <bind-users at lists.isc.org <mailto:bind-users at lists.isc.org>>
> Subject: RE: Sporadic Timeouts after upgrading to bind9.20
>  
>  
> From: Ondřej Surý <ondrej at isc.org <mailto:ondrej at isc.org>>
> Sent: Friday, September 6, 2024 4:08 PM
> To: Klaus Darilion <klaus.darilion at nic.at <mailto:klaus.darilion at nic.at>>
> Cc: Petr Špaček <pspacek at isc.org <mailto:pspacek at isc.org>>; bind-users at lists.isc.org <mailto:bind-users at lists.isc.org>; Klaus Darilion via bind-users <bind-users at lists.isc.org <mailto:bind-users at lists.isc.org>>
> Subject: Re: Sporadic Timeouts after upgrading to bind9.20
>  
> Are your running with options { reuseport no; };  ?
>  
> You might want to try that.
>  
> After setting reuseport no; (and UV_THREADPOOL_SIZE=12) I have not seen any timeouts anymore.
>  
> Anyway, this:
>  
> TID 8917:
> #0  0x00007b385aa6daa9     cds_lfht_destroy - /usr/lib/x86_64-linux-gnu/liburcu-cds.so.8.1.0
>  
> caught my eye. Are the zones you are hosting particularly large on GLUE?
>  
> I don’T know and I have not checked yet. One of the affected zones is .ch.  You could download the zone fromhttps://zonedata.switch.ch/ And they are using NSEC (not NSEC3 as I have written before)
>  
>  
>  
> Also if you have more eu-stack, can you confirm this is the pattern now?
>  
> After setting reuseport no; I do not have stack-traces any more. But if that would help you I can undo the workaround next week to collect traces.
>  
> Thanks
> Klaus
>  
>  
> -- 
> Visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe from this list
> 
> ISC funds the development of this software with paid support subscriptions. Contact us at https://www.isc.org/contact/ for more information.
> 
> 
> bind-users mailing list
> bind-users at lists.isc.org <mailto:bind-users at lists.isc.org>
> https://lists.isc.org/mailman/listinfo/bind-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20241205/cb9fa3d0/attachment-0001.htm>