[OpenAFS] replication over slower bandwidth

[muayad y]

--1982235058-1346517035-1329090354=:80274
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

> Are you seeing the same performance issues with the 1.6 version of=0A    =
'vos'?=A0=20
=A0 I tried it and it was terribly unstable=A0 on debian squeeze,
>=A0Can you confirm that commenting out the field and rebuilding=0A    rest=
ores the slow transmission speed with the 1.4 vos?r>=0A   =20
no, I did it again but there is no improvement, so uncommenting the maxPack=
etsize does not help, first, it want change the laignment cause it has size=
of=A0 8-bytes + it is not being referenced=A0 in the source tree. Probably =
that vos release occured while wan connection was functioning at max bandwi=
dth.=20

=0A    If the answer is 'yes' to the previous question, instead of removing=
=0A    the comment from maxPacketSize, do you see a similar performance=0A =
   benefit if you add a new field later in the structure?=A0 Perhaps=0A    =
after 'refCount' or 'securityIndex'?
=0A   =20
=0A    As Simon said, if there is a problem with how this structure is=0A  =
  being processed, we probably have a similar problem throughout the=0A    =
code base.=A0=A0 It is important for us to understand what the cause of=0A =
   the problem is.
=0A   =20
=0A   =20
=0A  =0A
--1982235058-1346517035-1329090354=:80274
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

<table cellspacing=3D"0" cellpadding=3D"0" border=3D"0" ><tr><td valign=3D"=
top" style=3D"font: inherit;">&gt; Are you seeing the same performance issu=
es with the 1.6 version of=0A    'vos'?&nbsp; <br>&nbsp; I tried it and it =
was terribly unstable&nbsp; on debian squeeze,<br>&gt;&nbsp;Can you confirm=
 that commenting out the field and rebuilding=0A    restores the slow trans=
mission speed with the 1.4 vos<span>?</span>r&gt;=0A    <br>no, I did it ag=
ain but there is no improvement, so unc<span>ommenting th</span>e maxPacket=
size does not help, first, it want change the laignment cause it has sizeof=
&nbsp; 8-bytes + it is not being referenced&nbsp; in the source tree. Proba=
bly that vos release occured while wan connection was functioning at max ba=
ndwidth. <br><br><blockquote style=3D"border-left: 2px solid rgb(16, 16, 25=
5); margin-left: 5px; padding-left: 5px;"><div id=3D"yiv1153180460"><div><b=
>=0A    If the answer is 'yes' to the previous question, instead of removin=
g=0A    the comment from maxPacketSize, do you see a similar performance=0A=
    benefit if you add a new field later in the structure?&nbsp; Perhaps=0A=
    after 'refCount' or 'securityIndex'?<br>=0A    <br>=0A    As Simon said=
, if there is a problem with how this structure is=0A    being processed, w=
e probably have a similar problem throughout the=0A    code base.&nbsp;&nbs=
p; It is important for us to understand what the cause of=0A    the problem=
 is.<br>=0A    <br>=0A    <br>=0A  </b></div>=0A</div></blockquote></td></t=
r></table>
--1982235058-1346517035-1329090354=:80274--