[OpenAFS] Re: freezes acessing /afs/.git
Markus Koeberl
spsc-sysadmin@mlist.tugraz.at
Thu, 14 Aug 2014 18:44:17 +0200
--Boundary-01=_hdO7TUuxlImY5tg
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
On Thursday 14 August 2014 18:11:29 Andrew Deason wrote:
> On Thu, 14 Aug 2014 10:06:31 +0200
> Markus Koeberl <markus.koeberl@tugraz.at> wrote:
>=20
> > On Thursday 14 August 2014 05:26:36 Andrew Deason wrote:
> > > This generally sounds reasonable to me. But I think you want to just
> > > require a dot if the name doesn't exist in the csdb at all, so you
> > > can just put it in the local csdb if you want to use it, without
> > > needing to alias it to a full cell name. (This may have been
> > > mentioned earlier, but I don't see it here.) I'd call such an option
> > > -afsdb-ndots, which would set the minimum number of dots (like
> > > resolv.conf's "option ndots").
> >=20
> > This would require changing the format of the CellServDB or did I get
> > something wrong?
>=20
> You can have a csdb entry that just lists a cell and doesn't list
> servers for it. Unless there's something else I said that sounded like
> changing the format...?
OK I got confused because in the csdb are domains and I was thinking of a c=
ase where there is no dot included.
Did you mean something like this:
$ test -d /afs/.git
1. first look if "git" is in csdb -> no match
2. add the search domain and look if it is in the csdb ("git.spsc.tugraz.at=
", "git.tugraz.at" -> both no match)
3. test if there are enough dots ("git" -> no match)
4. therefore not query the nameserver
$ test -d /afs/openafs.org
1. first look if "openafs.org" is in csdb -> no match
2. add the search domain and look if it is in the csdb ("openafs.org.spsc.t=
ugraz.at", "openafs.org.tugraz.at" -> both no match)
3. test if there are enough dots ("git" -> match)
4. query the nameserver for "openafs.org"
> > > A blacklist might help for a vast majority of cases, though. Or of
> > > course we could just improve the caching, so only the first lookup is
> > > slow... would that be good enough, if done "well"?
> >=20
> > A improved caching would be very nice for the bash completion.
>=20
> What I was trying to get at is "which is _more_ useful"? (Improved
> caching, or a blacklist/whitelist.) Or if the answer is "either one
> would make me happy", that would be good to know.
If I got it right now this would be a big win.
It will not help with bash completion for some cases e.g openafs.o + Tab.=20
Improved caching would help here because I think that people always write t=
he same number of chars before they press the Tab key. So hanging bash comp=
letion should not be that big problem.
regards
Markus K=F6berl
=2D-=20
Markus Koeberl
Graz University of Technology
Signal Processing and Speech Communication Laboratory
E-mail: markus.koeberl@tugraz.at
--Boundary-01=_hdO7TUuxlImY5tg
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0//EN" "http://www.w3.org/TR/REC-=
html40/strict.dtd">
<html><head><meta name=3D"qrichtext" content=3D"1" /><style type=3D"text/cs=
s">
p, li { white-space: pre-wrap; }
</style></head><body style=3D" font-family:'DejaVu Sans Mono'; font-size:9p=
t; font-weight:400; font-style:normal;">
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">On Thursday=
14 August 2014 18:11:29 Andrew Deason wrote:</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> On Thu=
, 14 Aug 2014 10:06:31 +0200</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> Markus=
Koeberl <markus.koeberl@tugraz.at> wrote:</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> </p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > O=
n Thursday 14 August 2014 05:26:36 Andrew Deason wrote:</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; This generally sounds reasonable to me. But I think you want to just</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; require a dot if the name doesn't exist in the csdb at all, so you</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; can just put it in the local csdb if you want to use it, without</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; needing to alias it to a full cell name. (This may have been</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; mentioned earlier, but I don't see it here.) I'd call such an option</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; -afsdb-ndots, which would set the minimum number of dots (like</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; resolv.conf's "option ndots").</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > <=
/p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > T=
his would require changing the format of the CellServDB or did I get</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > s=
omething wrong?</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> </p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> You ca=
n have a csdb entry that just lists a cell and doesn't list</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> server=
s for it. Unless there's something else I said that sounded like</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> changi=
ng the format...?</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">OK I got co=
nfused because in the csdb are domains and I was thinking of a case where t=
here is no dot included.</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Did you mea=
n something like this:</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">$ test -d /=
afs/.git</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">1. first lo=
ok if "git" is in csdb -> no match</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">2. add the =
search domain and look if it is in the csdb ("git.spsc.tugraz.at"=
, "git.tugraz.at" -> both no match)</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">3. test if =
there are enough dots ("git" -> no match)</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">4. therefor=
e not query the nameserver</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">$ test -d /=
afs/openafs.org</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">1. first lo=
ok if "openafs.org" is in csdb -> no match</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">2. add the =
search domain and look if it is in the csdb ("openafs.org.spsc.tugraz.=
at", "openafs.org.tugraz.at" -> both no match)</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">3. test if =
there are enough dots ("git" -> match)</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">4. query th=
e nameserver for "openafs.org"</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; A blacklist might help for a vast majority of cases, though. Or of</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; course we could just improve the caching, so only the first lookup is</=
p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > &=
gt; slow... would that be good enough, if done "well"?</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > <=
/p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> > A=
improved caching would be very nice for the bash completion.</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> </p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> What I=
was trying to get at is "which is _more_ useful"? (Improved</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> cachin=
g, or a blacklist/whitelist.) Or if the answer is "either one</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">> would =
make me happy", that would be good to know.</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">If I got it=
right now this would be a big win.</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">It will not=
help with bash completion for some cases e.g openafs.o + Tab. </p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Improved ca=
ching would help here because I think that people always write the same num=
ber of chars before they press the Tab key. So hanging bash completion shou=
ld not be that big problem.</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">regards</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Markus K=F6=
berl</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">-- </p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Markus Koeb=
erl</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Graz Univer=
sity of Technology</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">Signal Proc=
essing and Speech Communication Laboratory</p>
<p style=3D" margin-top:0px; margin-bottom:0px; margin-left:0px; margin-rig=
ht:0px; -qt-block-indent:0; text-indent:0px; -qt-user-state:0;">E-mail: mar=
kus.koeberl@tugraz.at</p>
<p style=3D"-qt-paragraph-type:empty; margin-top:0px; margin-bottom:0px; ma=
rgin-left:0px; margin-right:0px; -qt-block-indent:0; text-indent:0px; ">&nb=
sp;</p></body></html>
--Boundary-01=_hdO7TUuxlImY5tg--