Date: Sun, 18 Sep 2022 00:41:59 +0400 From: milky india <milkyindia@gmail.com> To: Alan Somers <asomers@freebsd.org> Cc: freebsd-fs <freebsd-fs@freebsd.org> Subject: Re: Need help with live system expansion (zfs+geli) Message-ID: <CACQUA7nM%2Bo90QsX_k%2B%2B4skBVJOo59mo0-P4yTyh-R7Dw=_Amdg@mail.gmail.com> In-Reply-To: <CAOtMX2g2YT%2Bd-p-MJdzAohCU3HqU9m%2B0U7ogBmCNUa4tgV3dLw@mail.gmail.com> References: <CACQUA7kC1E%2BHTEAFY=o%2BFpRcfmhJ-qPWuXseNvT5eJzK4Bcdpw@mail.gmail.com> <CAOtMX2g%2BGZ6SjQQPu-YWJhttYS3D7YhU-kAqMY4xHxXoe6rdVg@mail.gmail.com> <CACQUA7nD2Hrkvo5P7_9gyWkJwX8-ZhMAc3HdgrzTCK=XVqRMog@mail.gmail.com> <CAOtMX2iDg3H75Jp5-N5MgLAvBHKFs8UXvCVPBi_YQKMqYWsz6g@mail.gmail.com> <CACQUA7mzfvwRVhc00kGk3rZHKhFtMChZ983EJ0OBn90DJfaMUQ@mail.gmail.com> <CAOtMX2ijrHuhh8E=M%2BD=FKXq9c0tbOq-ZDNRQZ4n32U5=dEaVg@mail.gmail.com> <CACQUA7nTyT9hpguEbQ1LskHJxq1OUckswbpD__Fov-G9t8Myxg@mail.gmail.com> <CAOtMX2gLaur3c8TATuVeZdDNPoLcG%2B4BrTgOp=Ow3jAuStkKtA@mail.gmail.com> <CACQUA7=6YTN027oUa9xDPNc8Y1MS4rtVYwCxsc0HLpJ8G%2BC8yQ@mail.gmail.com> <CAOtMX2h-adnaupXrwH90HpepnnBpUZFY5OaF_e6a91vFCe8Vjw@mail.gmail.com> <CACQUA7kL7GJTBvxvSh6dnMZ9qzyOS--6S4xt7q5nSc2s=hKwdw@mail.gmail.com> <CAOtMX2hat0FbJWpJA-fQZGxuV_NPASMazTifrJe2WPK8a_%2BMLw@mail.gmail.com> <CACQUA7mYYwe_9vgEAkbaeAHsikTbhK5v9EswpTS41nnvDiVh9w@mail.gmail.com> <CAOtMX2g2YT%2Bd-p-MJdzAohCU3HqU9m%2B0U7ogBmCNUa4tgV3dLw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--00000000000023199805e8e58366 Content-Type: text/plain; charset="UTF-8" On Sun, Sep 18, 2022 at 12:30 AM Alan Somers <asomers@freebsd.org> wrote: > On Sat, Sep 17, 2022 at 2:25 PM milky india <milkyindia@gmail.com> wrote: > > > > > > > > On Sun, Sep 18, 2022 at 12:20 AM Alan Somers <asomers@freebsd.org> > wrote: > >> > >> On Sat, Sep 17, 2022 at 2:03 PM milky india <milkyindia@gmail.com> > wrote: > >> > > >> > > >> > > >> > On Sat, Sep 17, 2022 at 10:58 PM Alan Somers <asomers@freebsd.org> > wrote: > >> >> > >> >> On Sat, Sep 17, 2022 at 12:52 PM milky india <milkyindia@gmail.com> > wrote: > >> >> > > >> >> > > >> >> > > >> >> > On Sat, Sep 17, 2022 at 10:43 PM Alan Somers <asomers@freebsd.org> > wrote: > >> >> >> > >> >> >> On Sat, Sep 17, 2022 at 12:31 PM milky india < > milkyindia@gmail.com> wrote: > >> >> >> > > >> >> >> > Sorry about that, again - I'm not sure what you mean by > bottom-post vs top-post. > >> >> >> > > >> >> >> > Be that as it may - I read the geli man page. I was > specifically warned against using "geli resize" since it may not work as > expected > https://forums.FreeBSD.org/threads/how-to-extend-zfs-geli-encrypted-disk-space-not-showing.86447/post-581642 > >> >> >> > Is this advise wrong? > >> >> >> > > "The geli has autoresize flag which will handle the new > provider size after gpart resize command." > >> >> >> > followed by > >> >> >> > > You are right, no geli resize needed. > >> >> >> > > >> >> >> > What would be the correct sequence of commands to fix this - > Simply "geli resize" ? (the -s option seems to be additional, will it > figure it out without providing s?) > >> >> >> > > >> >> >> > On Sat, Sep 17, 2022 at 10:20 PM Alan Somers < > asomers@freebsd.org> wrote: > >> >> >> > >> >> >> top-posting is where you insert your reply above the previous > email. > >> >> >> Bottom posting is where you insert your reply below it, like I'm > >> >> >> doing. The forum user said that you shouldn't need to run "geli > >> >> >> resize" because the AUTORESIZE flag is on. But as you can see > from > >> >> >> your "geli list" output, it's actually off. So you need to run > "geli > >> >> >> resize". The "-s" flag should be unnecessary since your provider > is > >> >> >> already online. At any rate, you can try it both ways. You might > >> >> >> want to make a copy of /var/backups/ada0p3.eli, just in case you > make > >> >> >> a mistake. > >> >> > > >> >> > Thanks - I hope I am bottom posting this as was expected of me. > >> >> > > >> >> > So if I understand correctly the AUTORESIZE flag is present for > adaop2 and NOT for adap3 (which is the partition we are concerned about) - > hence the advice given to not use "geli resize" isn't applicable here. Am I > understanding this correctly? > >> >> > >> >> Yes > >> >> > >> >> > > >> >> > > So you need to run "geli resize" > >> >> > Is this the only command that I need to run to resize my geli > partition? > >> >> > >> >> Yes > >> >> > >> >> > > >> >> > > The "-s" flag should be unnecessary since your provider is > >> >> > already online. > >> >> > Ok thanks. > >> >> > > >> >> > >You might want to make a copy of /var/backups/ada0p3.eli, just in > case you make a mistake. > >> >> > Don't have the luxury of backup currently. > >> >> > >> >> Um, ok. I can't guess why you aren't able to do that, but it isn't > >> >> strictly necessary. > >> >> > >> >> > > >> >> > I suppose at the end of it - if this works - "geli list" would > reflect the size to be 458G? (vs 290G currently) > >> >> > And that's the output I can trust to have solved the issue - or is > there more to it? > >> >> > >> >> Yes. > >> > > >> > > The "-s" flag should be unnecessary since your provider is already > online. > >> > When I try to run "geli resize /dev/ada0p3.eli" it complains. So I > guess options s is must ? If yes - do I need to put in the exact size down > to bytes from the output of geli list ? Under "Mediasize" ? > >> > --------------------------------------- > >> > geli: Option 's' not specified. > >> > ---------------------------------------- > >> > >> Dude, it's easier just to try it, than to ask us. Go for it. The > >> worst case scenario if you get the argument wrong is that nothing > >> happens. > > > > > > > The worst case scenario if you get the argument wrong is that nothing > happens. > > I'm not very sure if I put in the wrong value of s (old size) then > "nothing happens" - I would imagine if I put in a size less than the > current size then possibly the data after that get's lost? Or if I put in a > size greater than the current size then there is a gap in the geli > partition? > > I sense some frustration in your reply - but I'm on a live system and > wouldn't want to possibly risk it going kaput at this last step. That's the > reason I'm trying to understand what is the best value of s for geli resize > and how to obtain it. > > The purpose of the "-s" argument is to tell geli where to find the old > label, if the provider isn't already attached. If you supply the > wrong argument, then geli won't be able to find the label, and thus > won't be able to do anything. It won't destroy any data, and it will > always automatically determine the size of the current provider. Of > course, if you're worried about losing data, you should always save a > copy of /var/backups/ada0p3.eli, as I suggested. > -Alan > >The purpose of the "-s" argument is to tell geli where to find the old > label, if the provider isn't already attached. If you supply the > wrong argument, then geli won't be able to find the label, and thus > won't be able to do anything. I think you're confusing the -s argument with something else - the man page says it's for the old size - that's why I was trying to figure it out and still haven't been able to quite yet. Although I suspect the "Mediasize" output for p3 from "geli list" is what the s value should be - but not very sure : ------------------------------------------------------------------------------------------ Additional options include: -s oldsize The size of the provider before it was resized. ------------------------------------------------------------------------------------------------ --00000000000023199805e8e58366 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable <div dir=3D"ltr"><div dir=3D"ltr"><br></div><br><div class=3D"gmail_quote">= <div dir=3D"ltr" class=3D"gmail_attr">On Sun, Sep 18, 2022 at 12:30 AM Alan= Somers <<a href=3D"mailto:asomers@freebsd.org">asomers@freebsd.org</a>&= gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px 0= px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Sa= t, Sep 17, 2022 at 2:25 PM milky india <<a href=3D"mailto:milkyindia@gma= il.com" target=3D"_blank">milkyindia@gmail.com</a>> wrote:<br> ><br> ><br> ><br> > On Sun, Sep 18, 2022 at 12:20 AM Alan Somers <<a href=3D"mailto:aso= mers@freebsd.org" target=3D"_blank">asomers@freebsd.org</a>> wrote:<br> >><br> >> On Sat, Sep 17, 2022 at 2:03 PM milky india <<a href=3D"mailto:= milkyindia@gmail.com" target=3D"_blank">milkyindia@gmail.com</a>> wrote:= <br> >> ><br> >> ><br> >> ><br> >> > On Sat, Sep 17, 2022 at 10:58 PM Alan Somers <<a href=3D"m= ailto:asomers@freebsd.org" target=3D"_blank">asomers@freebsd.org</a>> wr= ote:<br> >> >><br> >> >> On Sat, Sep 17, 2022 at 12:52 PM milky india <<a href= =3D"mailto:milkyindia@gmail.com" target=3D"_blank">milkyindia@gmail.com</a>= > wrote:<br> >> >> ><br> >> >> ><br> >> >> ><br> >> >> > On Sat, Sep 17, 2022 at 10:43 PM Alan Somers <<a = href=3D"mailto:asomers@freebsd.org" target=3D"_blank">asomers@freebsd.org</= a>> wrote:<br> >> >> >><br> >> >> >> On Sat, Sep 17, 2022 at 12:31 PM milky india <= ;<a href=3D"mailto:milkyindia@gmail.com" target=3D"_blank">milkyindia@gmail= .com</a>> wrote:<br> >> >> >> ><br> >> >> >> > Sorry about that, again - I'm not sure = what you mean by bottom-post vs top-post.<br> >> >> >> ><br> >> >> >> > Be that as it may - I read the geli man pag= e. I was specifically warned against using "geli resize" since it= may not work as expected <a href=3D"https://forums.FreeBSD.org/threads/how= -to-extend-zfs-geli-encrypted-disk-space-not-showing.86447/post-581642" rel= =3D"noreferrer" target=3D"_blank">https://forums.FreeBSD.org/threads/how-to= -extend-zfs-geli-encrypted-disk-space-not-showing.86447/post-581642</a><br> >> >> >> > Is this advise wrong?<br> >> >> >> > > "The geli has autoresize flag whi= ch will handle the new provider size after gpart resize command."<br> >> >> >> > followed by<br> >> >> >> > > You are right, no geli resize needed.<= br> >> >> >> ><br> >> >> >> > What would be the correct sequence of comma= nds to fix this - Simply "geli resize" ? (the -s option seems to = be additional, will it figure it out without providing s?)<br> >> >> >> ><br> >> >> >> > On Sat, Sep 17, 2022 at 10:20 PM Alan Somer= s <<a href=3D"mailto:asomers@freebsd.org" target=3D"_blank">asomers@free= bsd.org</a>> wrote:<br> >> >> >><br> >> >> >> top-posting is where you insert your reply above= the previous email.<br> >> >> >> Bottom posting is where you insert your reply be= low it, like I'm<br> >> >> >> doing.=C2=A0 The forum user said that you should= n't need to run "geli<br> >> >> >> resize" because the AUTORESIZE flag is on.= =C2=A0 But as you can see from<br> >> >> >> your "geli list" output, it's actu= ally off.=C2=A0 So you need to run "geli<br> >> >> >> resize".=C2=A0 The "-s" flag shou= ld be unnecessary since your provider is<br> >> >> >> already online.=C2=A0 At any rate, you can try i= t both ways.=C2=A0 You might<br> >> >> >> want to make a copy of /var/backups/ada0p3.eli, = just in case you make<br> >> >> >> a mistake.<br> >> >> ><br> >> >> > Thanks - I hope I am bottom posting this as was expe= cted of me.<br> >> >> ><br> >> >> > So if I understand correctly the AUTORESIZE flag is = present for adaop2 and NOT for adap3 (which is the partition we are concern= ed about) - hence the advice given to not use "geli resize" isn&#= 39;t applicable here. Am I understanding this correctly?<br> >> >><br> >> >> Yes<br> >> >><br> >> >> ><br> >> >> > > So you need to run "geli resize"<br> >> >> > Is this the only command that I need to run to resiz= e my geli partition?<br> >> >><br> >> >> Yes<br> >> >><br> >> >> ><br> >> >> > > The "-s" flag should be unnecessary s= ince your provider is<br> >> >> > already online.<br> >> >> > Ok thanks.<br> >> >> ><br> >> >> > >You might want to make a copy of /var/backups/ad= a0p3.eli, just in case you make a mistake.<br> >> >> > Don't have the luxury of backup currently.<br> >> >><br> >> >> Um, ok.=C2=A0 I can't guess why you aren't able t= o do that, but it isn't<br> >> >> strictly necessary.<br> >> >><br> >> >> ><br> >> >> > I suppose at the end of it - if this works - "g= eli list" would reflect the size to be 458G? (vs 290G currently)<br> >> >> > And that's the output I can trust to have solved= the issue - or is there more to it?<br> >> >><br> >> >> Yes.<br> >> ><br> >> > > The "-s" flag should be unnecessary since your= provider is already online.<br> >> > When I try to run "geli resize /dev/ada0p3.eli" it = complains. So I guess options s is must ? If yes - do I need to put in the = exact size down to bytes from the output of geli list ? Under "Mediasi= ze" ?<br> >> > ---------------------------------------<br> >> > geli: Option 's' not specified.<br> >> > ----------------------------------------<br> >><br> >> Dude, it's easier just to try it, than to ask us.=C2=A0 Go for= it.=C2=A0 The<br> >> worst case scenario if you get the argument wrong is that nothing<= br> >> happens.<br> ><br> ><br> > > The worst case scenario if you get the argument wrong is that not= hing happens.<br> > I'm not very sure if I put in the wrong value of s (old size) then= "nothing happens" - I would imagine if I put in a size less than= the current size then possibly the data after that get's lost? Or if I= put in a size greater than the current size then there is a gap in the gel= i partition?<br> > I sense some frustration in your reply - but I'm on a live system = and wouldn't want to possibly risk it going kaput at this last step. Th= at's the reason I'm trying to understand what is the best value of = s for geli resize and how to obtain it.<br> <br> The purpose of the "-s" argument is to tell geli where to find th= e old<br> label, if the provider isn't already attached.=C2=A0 If you supply the<= br> wrong argument, then geli won't be able to find the label, and thus<br> won't be able to do anything.=C2=A0 It won't destroy any data, and = it will<br> always automatically determine the size of the current provider.=C2=A0 Of<b= r> course, if you're worried about losing data, you should always save a<b= r> copy of /var/backups/ada0p3.eli, as I suggested.<br> -Alan<br></blockquote><div><br></div><div>=C2=A0>The purpose of the &quo= t;-s" argument is to tell geli where to find the old</div>> label, = if the provider isn't already attached.=C2=A0 If you supply the<br>>= wrong argument, then geli won't be able to find the label, and thus<br= >> won't be able to do anything.</div><div class=3D"gmail_quote"><br= ></div><div class=3D"gmail_quote">I think you're confusing the -s argum= ent with something else - the man page says it's for the old size - tha= t's why I was trying to figure it out and still haven't been able t= o quite yet. Although I suspect the "Mediasize" output for p3 fro= m "geli list" is what the s value should be - but not very sure := <br>-----------------------------------------------------------------------= -------------------</div><div class=3D"gmail_quote">Additional options incl= ude:<br><br>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 -s oldsize =C2=A0 =C2=A0 =C2=A0 =C2=A0The size of the provider befor= e it was resized.<br></div><div class=3D"gmail_quote">---------------------= ---------------------------------------------------------------------------= </div><div class=3D"gmail_quote"><br></div></div> --00000000000023199805e8e58366--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CACQUA7nM%2Bo90QsX_k%2B%2B4skBVJOo59mo0-P4yTyh-R7Dw=_Amdg>