SilverTip257 wrote:

On Fri, Feb 15, 2013 at 2:43 PM, m.roth@5-cent.us wrote:

...

SilverTip257 wrote:

...

Is anybody here monitoring Dell SAS6/iR RAID health? I have a mix of CentOS 5.9 and 6.3 installs on the PE1950s.

<snip> Um, um, er, uh... how old are the PE1950's? Let me urge you in the *strongest* terms to get your PO's out for replacements, yesterday if possible.

I hear you. Frankly, I have to work with what hardware I'm given. Even if that means making a shoe string into an Ethernet cable.

...

We had a lot of them. Then, about two years ago, they started dying - inside of a month, we had three? four? die, and it continued. What was happening was the RAID controller - if you really want, I can go look up, in old emails, the error number that showed on the LCD screen. When

that

...

happened, it was gone, dead.

I will admit to being amazed at the quality control, that they *all* started dying in such a short timespan.

<snip> Please feel free to pass what I said along to your management. You might consider coming up with a disaster plan for how you'd deal with two of these servers dying within a month, and present that to them when you talk to them....

mark

On Fri, Feb 15, 2013 at 3:16 PM, Todor Petkov zakk@online.bg wrote:

On 15/02/2013 09:08 PM, SilverTip257 wrote:

...

Is anybody here monitoring Dell SAS6/iR RAID health? I have a mix of CentOS 5.9 and 6.3 installs on the PE1950s.

I'm finding drivers and Megaraid Storage Manager downloads via LSI's site, but I'd prefer not to install a bunch of crapware. All I need is the proper LSI daemon which exports info via SNMP.

I've used check_sasraid_megaraid Nagios script [0] in the past for PERC5i and PERC6i controllers in Dell PE2950s. For 2950s I've only had to install sas_snmp-3.11-0003.i386.rpm from LSI to get rolling.

I've tried installing parts from the MSM zip/tarball mess to install only the lsi_mrdsnmpd daemon (lsi_mrdsnmpagent).

PERC5i = LSI 8048e (megaraid) PERC6i = LSI SAS 1078 (megaraid) SAS6/iR = LSI SAS 1068e (mptfusion)

[0] http://wleibzon.bol.ucla.edu/nagios/ [1] http://lists.centos.org/pipermail/centos/2009-November/085215.html

Thanks for any nudges in the right direction.

You can try megasasctl from this package http://sourceforge.net/projects/megactl/ and do some script, which is used by snmpd to check the output.

Thanks - I'll look into it.

Not as nice as the lsi daemon though and having to re-author a Nagios script.

Regards,

---

CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos