Hi -
Once again, Netapp tech support seems to be taking their sweet time about responding to my support email, so I'm going to folks that actually respond...
Got this in the logs of one of my 720s. Is this telling me I've got bad RAM?
There was an error in the J28 DIMM that was corrected, a
couple of times.
It hasn't recurred since that date, but still worries me.
Since it hasn't happened again, I wouldn't worry about it. The timing seems to indicate that perhaps something was going on unusual at that time, perhaps a raid scrub or a heavy load from cron or somesuch? It is not unusual to see more errors during times of heavier stress.
You might want to consider reseating that DIMM just to make sure it has a good connection. Then I wouldn't worry unless you start seeing a LOT more errors. But support may tell you different.
Bruce
Thanks, Bruce.
Actually, it repeated several hundred times that night, never since. First time for that error in over a year in owning it, which is why it made me twitchy. Cron-wise, nothing interesting changed. Scrubbing has been going on forever.
Yikes, I hate to play this loose with it, but it will be mirrored to another 720 in a few weeks, so hopefully it won't matter.
We're extremely dependent on this filer. I know we're a minor player in Netapp's realm ( we own 4 720s ), but the lack of response for such a critical piece of machinery for me makes me wonder. Guess I won't blow off the next EMC call that is due to hit voice mail in a couple of weeks.
Thanks again for the note, Bruce.
-j
Actually, it repeated several hundred times that night, never since. First time for that error in over a year in owning it, which is why it made me twitchy. Cron-wise, nothing interesting changed. Scrubbing has been going on forever.
Oh, sorry, I must have missed that fact in your first email. Strange though, that it would happen so much that night and then not again. Makes me wonder if a bug didn't crawl in the filer. :)
Yikes, I hate to play this loose with it, but it will be mirrored to another 720 in a few weeks, so hopefully it won't matter.
We're extremely dependent on this filer. I know we're a minor player in Netapp's realm ( we own 4 720s ), but the lack of response for such a critical piece of machinery for me makes me wonder. Guess I won't blow off the next EMC call that is due to hit voice mail in a couple of weeks.
My guess is a lack of response means it is either very unusual, not very serious, or both. You're really not in any "danger" in that if it fails again, and keeps failing, you'll just have to pull the DIMM and replace it. But that hasn't happened. Is it more likely to fail? I would have thought yes, but you haven't seen any errors since, so....
I would probably take the filer down and reseat the DIMM and the others as well just to be safe. You can also do nothing, or your other option is to replace it and give the other one back to Netapp. Of course, you never know if the replacement memory could be faulty...
Bruce
On Sat, 6 May 2000, Jamie A. Lawrence wrote:
since. First time for that error in over a year in owning it, which is why it made me twitchy. Cron-wise, nothing interesting changed. Scrubbing has been going on forever.
Yikes, I hate to play this loose with it, but it will be mirrored to another 720 in a few weeks, so hopefully it won't matter.
In terms of comfort level, I've had a memory module or two spit out one or two minor errors (not hundreds, however) and after talking with NetApp support, just left the modules in and left them alone. Other times, I've asked NetApp support to send me replacement parts to "calm my nerves" and they seem willing to be on the safe side, not the sorry side, and RMA replacements.
Basically, if you are not satisfied with their lack of response (as Bruce said, they may deem it a minor problem or just a notification really), then call them up and ask for replacement. Isn't that what customer satisfaction is all about?
Until next time...
The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill@mathworks.com http://www.mathworks.com ---
-
Bruce Sterling Woodcock -
Jamie A. Lawrence -
Todd C. Merrill