Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 3 Jun 2019 23:13:35 +0000
From:      Rick Macklem <rmacklem@uoguelph.ca>
To:        Alan Somers <asomers@freebsd.org>, Peter Eriksson <pen@lysator.liu.se>
Cc:        "freebsd-fs@freebsd.org" <freebsd-fs@freebsd.org>
Subject:   Re: RFC: should an incremental reload of exports for mountd be optional?
Message-ID:  <YQXPR01MB31289FF0AC8633D372C814A2DD140@YQXPR01MB3128.CANPRD01.PROD.OUTLOOK.COM>
In-Reply-To: <CAOtMX2hs-aQgEZaCQjNdMHuL8qxJTqnnvNQ%2BysRcpFFTEtDb1A@mail.gmail.com>
References:  <QB1PR01MB2643E5186098E2AF14934866DD190@QB1PR01MB2643.CANPRD01.PROD.OUTLOOK.COM> <D685FA4E-0F35-44F8-B341-CF59122E10B9@lysator.liu.se>, <CAOtMX2hs-aQgEZaCQjNdMHuL8qxJTqnnvNQ%2BysRcpFFTEtDb1A@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Thanks everyone for your comments. I just committed the patch with
the incremental reload of exports always enabled.
If people run into problems, I can add some "backdoor" way to disable
it, as suggested by asomers@.

A restart of the daemon will always do a full reload to work around any
reload failures.

rick

________________________________________
From: Alan Somers <asomers@freebsd.org>
Sent: Monday, June 3, 2019 6:28:51 PM
To: Peter Eriksson
Cc: Rick Macklem; freebsd-fs@freebsd.org
Subject: Re: RFC: should an incremental reload of exports for mountd be opt=
ional?

I think that it should be always enabled.  If the only reason not to
is a fear of bugs in the new method, then the old method can be
selected by some back-doorish workaround, like an environment
variable.
-Alan

On Mon, Jun 3, 2019 at 4:16 PM Peter Eriksson <pen@lysator.liu.se> wrote:
>
> My vote goes to:
>
> Enabled by default (with an option to disable it in case problems pop up)=
.
>
> - Peter
>
>
> > On 31 May 2019, at 23:50, Rick Macklem <rmacklem@uoguelph.ca> wrote:
> >
> > Hi,
> >
> > I have created a patch for PR#237860 that improves the time it takes to=
 reload
> > the exports file(s) for an NFS file server from seconds->milliseconds b=
y only doing
> > system calls for the changes to the exports file(s). (Assumes each relo=
ad includes
> > changes to only a small fraction of the entries.)
> >
> > I made it a non-default option of "-I", but asomers@ asked why it neede=
d to be
> > optional.
> >
> > Basically, I made it optional because I felt the patch might still be b=
uggy and not
> > handle some case of an exports file change that I missed.
> >
> > However, I can see the argument of making it non-optional (or optional =
but enabled
> > by default) so that sysadmins don't need to add "-I" to avoid long peri=
ods where the
> > nfsd threads are suspended.
> > If it is broken, the deamon would need to be restarted to do a full rel=
oad, but that
> > would also be the case if a sysadmin specified the "-I" option.
> >
> > So, the question is... "do you think this should be an option or just a=
lways enabled?".
> >
> > Thanks , rick
> > _______________________________________________
> > freebsd-fs@freebsd.org mailing list
> > https://lists.freebsd.org/mailman/listinfo/freebsd-fs
> > To unsubscribe, send any mail to "freebsd-fs-unsubscribe@freebsd.org"
>
> _______________________________________________
> freebsd-fs@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-fs
> To unsubscribe, send any mail to "freebsd-fs-unsubscribe@freebsd.org"



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?YQXPR01MB31289FF0AC8633D372C814A2DD140>