Will SPS Test run if SPS B failed ?

I did raise SR with EMC 1 week back and they replied stating that Clariion has self-test mechanism which when scheduled SPS fails and when the test completes both the SPS comes online. If the SPS cabling status is in valid for longer periods of time your actual h/w might be failure. Please contact your H/W vendor for replacement.

it is probally the normal behavior since i dont think it would go ahead with the test because the system does not know if the SPS on the B side will keep the SP up.  it would risk a SP shutdown if it tried to run the test.

We are pushing DELL to replace the SPS.  When it has been replaced, we will test the SPS TEST function and will keep you posted.

Thanks

Hi,

Thanks for your advice.

Yes, the SPS test is sheduled to be conducted on Sundays.

However, the Faulty sign doesn't disappear after the test.  We have talked to DELL for a number of times and they have a lot of excuses for agruing that it was not a Hardware Problem though we have 4 hours support contract with them.

Thanks again

Tony, when the SPS is replaced it should go through a charge cycle, then a test cycle. You shouldnt need to  force a test.

jim

We get the Email notification this morning saying that SPS Test hasn't been conducted for 2 weeks.

Severity Error Host CX4_SPA
Storage Array CKM00088800012 SPA Device Enclosure 0 SPS A
Description Weekly SPS Test has not been executed for two consecutive weeks. Please verify SPSs and their related components or contact your service provider.

Regards,
Tony

Tony,  it will not do the test untill the fault is cleared whether it is a bad SPS or a bad SPS sense cable.  You said you got it from Dell correct?  if you are getting the run around, you should talk to your TAM and get it escalated.

Hi,

I have the same problem with CX4-480 Flare: 04.30.000.5.517.

The Vendor changed 2 times the SPS and the serial cable between SPS and SP.

No results yet.

  (Enclosure SPE : Faulted; Enclosure SPE SPS B : Faulted)

Enclosure SPE Power A0 State: Present

Enclosure SPE Power A1 State: Present

Enclosure SPE Power B0 State: Present

Enclosure SPE Power B1 State: Present

Enclosure SPE SPS A State:  Present

Enclosure SPE SPS B State:  Present

Enclosure SPE SPS A Cabling State: Valid

Enclosure SPE SPS B Cabling State: Cabling Status is unknown

This evening when I tried to "force" a test, an event code 0x916 was generated by the array:

"Weekly SPS Test has not been executed for two consecutive weeks. Please verify SPSs and their related components or contact your service provider."

Follow by the event code 0x62a:

"1.  Periodic SPS test time.  2.  No action needed unless this message is seen more frequently than the periodic SPS Test is setup to run."

SpCollects was generated and sent to EMC.

I wait for a solution...

BR,

AlexB

Hi,

I read about 0x62a:

0x62A The timer used to schedule the periodic SPS Test has expired. No action needed unless this message is seen more

frequently than the periodic SPS Test is setup to run.

Can anybody give more details about this timer that has expired?

If no other solution, in the next Maintenance Window I will restart the SP (one by one) and I guess all will be fain again.

SP has now 202 days uptime.

BR,

AlexB

Hi,

EMC Primus: emc212449

"

Fix:

The software issues have been fixed through several releases and are as follows:

1. Bad SPS Sense Cable – Being worked by manufacturing.
2. Cabling Status Unknown – An unexpected status response from the 1.2KW SPS can leave the SPS stuck in a faulted state. This issue has been addressed in the initial release of Release 29, Release 28.707, Release 23.707 and Q1’09 Release 26.031 patch.
3. Unknown Config reported – The array can get stuck in a state where one or both SPS’s are marked invalid due to unknown cabling. This is caused by an unexpected response from the 1.2KW SPS immediately after a BATTEST command was sent. This issue has been addressed in the Q2 2010 patch release of R29 patch 012.

Workaround:

Use the following workaround in order of increasing impact for immediate mitigation to the problem:

1. Pull sense cable inspect for pushed pins, replace if necessary. If not, hold out for at least three seconds then re-insert. Perform the action on the peer SP only if necessary.
2. Reboot the SP on the side that is reporting the error.
3. Reboot the other SP.
4. Cold power cycle the xPE enclosure.

After performing above workaround, verify from the GUI that the SPS is no longer reported as faulted.

Warning! Under NO circumstances should the SPS cables be swapped around if one of the SPS is working correctly. This could lead to write cache being DISABLED.

"

Reboot is the official way...

BR,

AlexB

In this particular case, error Cabling Status Unknown, the failed controller B, the controller of battery also assume the cache controller B? Ie my cache would not be turned off because the batteries are mirrored? Only there is a risk of data loss if the error existed in dua enclosures SPE SPS?

There should be no data loss with a battery issue. If the SP's sense there is a problem with the batteries then the Write cache will disable and flush all data to disk or vault.

The loss of one battery will not disable Write cache as one battery is sufficient to flush the cache in case of a power failure. See the following KB article for more information:

https://support.emc.com/kb/13711

glen

I have the same situation。

The SPS B of a CX4 SAN reported as faulty - Cabling Status is unkown，

The SPS A is still working but we find that the scheduled SPS Test is not running，and have the following alert:

SPS Test hasn't been conducted for 2 weeks.

Severity Error Host CX4_SPA
Storage Array CKM000888000XX SPA Device Enclosure 0 SPS A
Description Weekly SPS Test has not been executed for two consecutive weeks. Please verify SPSs and their related components or contact your service provider.

we have change the SPS BATTERY TEST TIME, why SPA have the alert?

If SPB is still showing as a fault then the test will not run. Have you verified that the SPB is not faulted?

glen