From: "Zhenrong (Daniel) Gu" <zrgu@stu.xjtu.edu.cn>
To: Julian Anastasov <ja@ssi.bg>
Cc: horms@verge.net.au, lvs-devel@vger.kernel.org,
netdev@vger.kernel.org, Xing Feng <xingfeng@stu.xjtu.edu.cn>,
Peng Zhang <p-zhang@xjtu.edu.cn>
Subject: Re: [BUG] ipvs: new connection on a reused port goes to the previous real server during conn_tab resize
Date: Thu, 19 Nov 2026 06:39:20 -0500 [thread overview]
Message-ID: <av7f6F123V2WDd_u@fxxowodeMacBook-Pro.local> (raw)
In-Reply-To: <d876f5ef-982a-0ddb-5e75-3434453c2ccf@ssi.bg>
Hello Julian,
On Wed, Nov 18, 2026 at 05:26:40PM +0200, Julian Anastasov wrote:
>
> We prefer not to unhash such connections because they will
> go out of our control, eg. we may want to delete them immediately
> on netns cleanup. May be we should add some check to make them
> invisible to packets, such as adding new bit in the cp->flags,
> not depending on the cp->timeout being 0 or not. I can prepare
> some patch to fix this problem.
Thanks, a flag that hides the entry from packet lookups makes sense
to us, and it also keeps ip_vs_conn_put() from re-arming the old
entry's timer. We would be glad to test the patch: we can run the
reproducer (force, grow and noresize modes) on 7.3.3 and report back.
> They end up in the same table but in reverse order. The
> old table is freed after RCU grace period, so it should not matter
> which table is checked first.
Understood, thanks for the explanation.
Best regards,
Zhenrong (Daniel) Gu
next prev parent reply other threads:[~2026-11-19 11:39 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-17 15:38 [BUG] ipvs: new connection on a reused port goes to the previous real server during conn_tab resize Zhenrong (Daniel) Gu
2026-11-18 15:26 ` Julian Anastasov
2026-11-19 11:39 ` Zhenrong (Daniel) Gu [this message]
2026-11-20 4:55 ` Julian Anastasov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=av7f6F123V2WDd_u@fxxowodeMacBook-Pro.local \
--to=zrgu@stu.xjtu.edu.cn \
--cc=horms@verge.net.au \
--cc=ja@ssi.bg \
--cc=lvs-devel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=p-zhang@xjtu.edu.cn \
--cc=xingfeng@stu.xjtu.edu.cn \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.