JimiMuller
1 Copper

Diagnostic Log Question

I'm operating an EQL PS6100 storage array in a disconnected site where items must be (absolutely no exceptions) scrutinized before they are sent out. Diagnostics from the array are encrypted and are not human-readable. I've been able to work with Support without diagnostic logs but am now running into a recurring issue.

I'd like to get these logs to Dell in an effort to identify a root cause. Is there a way to generate unencrypted logs or a certificate of sorts offered by Dell which indicates the data which is found in the logs themselves?

Thanks for any input.

0 Kudos

RE: Diagnostic Log Question

Hello, 

 There is only an abbreviated diag that isn't encrypted.  That's useful for drive or controller failures.  Not sufficient for RCA triage. 

 The data is the output of various commands, that don't have access to actual data.  But I do understand the need for caution. 

 I'm not sure what you mean by a "certificate".   Do mean something written certifying that no user data is gathered? 

 Regards, 

Don 

 

Social Media and Community Professional
#IWork4Dell
Get Support on Twitter - @dellcarespro

JimiMuller
1 Copper

RE: Diagnostic Log Question

Hi, Don:
That's exactly what I mean. Is there a document which identifies what data is actually collected? Essentially, I need to be able to verify that there is no user data or specific bits of information about the data in the array itself within the logs before they can leave the facility.

If there is nothing like that available, how would you recommend I proceed with identifying a root cause of an issue? Would an on-site representative have equipment with which to view the logs?

Cheers,
James

0 Kudos

RE: Diagnostic Log Question

Hello, 

 There is no document like that, that I am aware of.   Doesn't mean one could not be written.  

Short version is the diag logs are the output of various proprietary commands that dump the logs, the group DB, info about cache, H/W status, network status,.drive history, etc.   As a block device the array doesn't  have access to the flies.  Individual blocks are arraigned in 15MB pages.  Each page could have different parts of a volume on it.  Data on the pages isn't sequentially used.  So even if the controller could get to the data, it's like unscrambling an egg to the OS EQL uses to boot the controller.  The software handling network, iSCSI and all admin operations are handled by EQL code.   Only HTTP:/SSH/Telenet/ftp are handled by the default OS.  

 Re: RCA.  Without the complete diags I would say it's impossible to determine RCA.  Additionally, the logs are circular, so the longer you wait, the more likely some data will be rolled out of the logs. 

ReSmiley Surprisednsite.  No, the field technicians are training to replace the failed H/W not triage. 

Regards, 

Don 

Social Media and Community Professional
#IWork4Dell
Get Support on Twitter - @dellcarespro

0 Kudos