Re: Weird yum/kernel problem

Bob McKay <[email protected]> Sun, 1 Feb 2015 19:01:59 +0900
Newsgroups gmane.linux.rpm.yum
Message-ID <[email protected]>
--===============6047946523326995539==
Content-Type: multipart/alternative; boundary="Apple-Mail=_570B55C1-2517-448C-AECC-E03FB5535FBF"


--Apple-Mail=_570B55C1-2517-448C-AECC-E03FB5535FBF
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

Hi Brian, and thank you for your comments - my apologies for being so =
long replying, I seem to have set up the mailing list wrong so that I=92m =
not receiving responses. Not sure how to fix it, so I guess I=92ll just =
need to check the archives regularly.

We bought the soft controller for cost reasons: we=92re a small =
university lab, and we just couln=92t afford hardware RAID controllers =
at the time. That argument applies all the more strongly now, as for =
various reasons the system only needs to survive another year or so. So =
unfortunately we=92re stuck with a software controller. I could just =
skip upgrading the system for that time, but we=92re receiving a huge =
number of attempted break-ins every day, so I prefer not to take =
security risks.

Thank you for the ubuntu link. I=92ve gone through it carefully, but I =
coulnd=92t find anything useful there (I don=92t want to use dkms =
directly, because failures under dims would be even harder to fix - at =
least if there is a failure building the module manually, I get to see =
it at the time, which means I can reset the kernel default to the =
preceding kernel, reboot, remove the latest kernel and rebuild again - =
and I can do this remotely (the system is in a lab a fair way form my =
home/work). Relying on DKMS in this circumstance, where builds fail on a =
regular basis, sounds a bit too dangerous.

    Thanks and Best Wishes
    Bob

   =20



Bob,
Some questions & recommendations:
Why not use a supported RAID controller for a production server?

Have you tried using DKMS?  I found a guide for Ubuntu and a lot of what =
it says applies to Fedora:

https://help.ubuntu.com/community/RocketRaid

If I were the sysadmin of your box and it needed to be online as much as =
possible, I would use a LSI RAID controller.  There are others out there =
which work, but I=92m most familiar with LSI.  Many vendors integrate =
their RAID controllers into their own RAID cards (Dell PERC, for =
example).  I would most definitely stay away from RAID controllers where =
the driver is not integrated into the mainstream kernel.

Just my 2 cents.

/Brian/
--
       Brian Long                             |       |
       Research Triangle Park, NC         . | | | . | | | .
                                              '       '
                                              C I S C O


--Apple-Mail=_570B55C1-2517-448C-AECC-E03FB5535FBF
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=windows-1252

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dwindows-1252"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: after-white-space;">Hi =
Brian, and thank you for your comments - my apologies for being so long =
replying, I seem to have set up the mailing list wrong so that I=92m not =
receiving responses. Not sure how to fix it, so I guess I=92ll just need =
to check the archives regularly.<div><br></div><div>We bought the soft =
controller for cost reasons: we=92re a small university lab, and we just =
couln=92t afford hardware RAID controllers at the time. That argument =
applies all the more strongly now, as for various reasons the system =
only needs to survive another year or so. So unfortunately we=92re stuck =
with a software controller. I could just skip upgrading the system for =
that time, but we=92re receiving a huge number of attempted break-ins =
every day, so I prefer not to take security =
risks.</div><div><br></div><div>Thank you for the ubuntu link. I=92ve =
gone through it carefully, but I coulnd=92t find anything useful there =
(I don=92t want to use dkms directly, because failures under dims would =
be even harder to fix - at least if there is a failure building the =
module manually, I get to see it at the time, which means I can reset =
the kernel default to the preceding kernel, reboot, remove the latest =
kernel and rebuild again - and I can do this remotely (the system is in =
a lab a fair way form my home/work). Relying on DKMS in this =
circumstance, where builds fail on a regular basis, sounds a bit too =
dangerous.</div><div><br></div><div>&nbsp; &nbsp; Thanks and Best =
Wishes</div><div>&nbsp; &nbsp; Bob</div><div><br></div><div>&nbsp; =
&nbsp;&nbsp;</div><div><br></div><div><br></div><div><br></div><div><hr> =
=20

<pre>Bob,
Some questions &amp; recommendations:
Why not use a supported RAID controller for a production server?

Have you tried using DKMS?  I found a guide for Ubuntu and a lot of what =
it says applies to Fedora:

<a =
href=3D"https://help.ubuntu.com/community/RocketRaid">https://help.ubuntu.=
com/community/RocketRaid</a>

If I were the sysadmin of your box and it needed to be online as much as =
possible, I would use a LSI RAID controller.  There are others out there =
which work, but I=92m most familiar with LSI.  Many vendors integrate =
their RAID controllers into their own RAID cards (Dell PERC, for =
example).  I would most definitely stay away from RAID controllers where =
the driver is not integrated into the mainstream kernel.

Just my 2 cents.

/Brian/
--
       Brian Long                             |       |
       Research Triangle Park, NC         . | | | . | | | .
                                              '       '
                                              C I S C =
O</pre><div><br></div></div></body></html>=

--Apple-Mail=_570B55C1-2517-448C-AECC-E03FB5535FBF--

--===============6047946523326995539==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Yum mailing list
[email protected]
http://lists.baseurl.org/mailman/listinfo/yum

--===============6047946523326995539==--