Unsolved
This post is more than 5 years old
10 Posts
0
4223
Fail to connect after several hours
I had made a post a while ago about a specific SDK error, listed here:
03 Sep 12, 20:26:29,324 [Thread-65] ERROR CenteraUtils- FP Library Exception: Cannot Open Connection
com.filepool.fplibrary.FPLibraryException: no primary cluster found
at com.filepool.fplibrary.FPLibraryException. (Unknown Source)
This is the previous post.
The last time that was due to a firewall issue, which we resolved. Currently the code accessing Centera works for 2-5 hours, and then it spits out this error until we restart the VM the code is running from. I'm not entirely convince this is an issue related to EMC and not a netowrk, but I'm being asked to check all possibilities at once.
Thanks
Tom
Satish_Kutty
216 Posts
0
September 5th, 2012 06:00
Hi Tom,
Refer to the Primus Solution ID ""emc138065" for more information on the access issues on Centera SDK.
Regards
Satish.N.Kutty
Satish_Kutty
216 Posts
1
September 5th, 2012 06:00
Hi Tom,
You'll be able to locate the Solution on Powerlink by doing search a for the Solution ID under the "Search Support" tab.
Regards
Satish.N.Kutty
Noxx
10 Posts
0
September 5th, 2012 06:00
I'm sorry, I don't know where to look for this. A search on the site or on google just gives me results that are posts of you saying to check "emc138065".
EMCDennis
124 Posts
1
September 7th, 2012 09:00
While this may very well be a network issue, it is possible that the access node you are referencing within the application connection string is down.
Make sure you have 2 or more IP addresses listed within the connection string for various access nodes on this single Centera. The API will use the first IP address listed and then use the second address should the first become unreachable.
Since you state that restarting the VM fixes this, this will not likely fix this issue but is worth mentioning
Dennis
Noxx
10 Posts
0
September 7th, 2012 10:00
We had various failures with our backup Centera device the other day and had to replace a Node and some other hardware from what I've been told...The problem occurred once since then which returned a -10101 error code. We're waiting for replication to finish and then we'll see if we still have the issue.
Thanks for the help so far.
Satish_Kutty
216 Posts
0
September 10th, 2012 05:00
Hi,
Thanks for the update......
Regards
Satish.N.Kutty
Satish_Kutty
216 Posts
0
September 12th, 2012 06:00
Hi,
Let me know if replication is completed and were you able to fix the issue...
Regards
Satish.N.Kutty
N0xx1
6 Posts
0
September 13th, 2012 06:00
It finally broke again last night but in the process of keeping the log file small it was cleared and after that stopped working, so we're waiting again...If we restart before checking the log it gets cleared and the person who restarted it last time didn't know that.
Satish_Kutty
216 Posts
0
September 13th, 2012 10:00
HI,
Thanks for the update...
Regards
Satish.N.Kutty
Phukon
337 Posts
1
September 17th, 2012 01:00
It looks like a an error occurred in Network Socket layer. When content address gets written to the Replica cluster even though the blobs associated with the metadata never make it across. I think you can Request to Run IC (Integrity checker) to get a clearer picture.
Regards,
Phukon.
N0xx1
6 Posts
0
September 18th, 2012 07:00
We solved all of our hardware and replication issues and had one more failure since then. We have run nearly a week now without any problems so we're monitoring a bit more but we're fairly certain the problem has been resolved.
EMC support suggested restarting each network access node one by one because sometimes that is needed after resolving hardware issues.
Thanks everyone.
Satish_Kutty
216 Posts
0
September 18th, 2012 07:00
Hi,
Thanks for the update. Please mark the closest solution as Correct or Helpful answer if possible.
Regards
Satish.N.Kutty