| View previous topic :: View next topic |
| Author |
Message |
Nonapeptide
Joined: 08 Sep 2012
Posts: 40
Location: Scottsdale, Arizona
|
 Posted: Sat Feb 02, 2013 1:10 am Post subject: Clusters of "Nothing received for 3 seconds" |
 |
|
While perusing the error.log file, I will occasionally see clusters of warnings that a small portion of mount point has not received anything for 3 seconds. For example, in an eleven second span of the logs I saw this:
| Quote: |
[2013-01-31 17:13:19] WARN source/source_read Nothing received on /mount1 for 3 seconds
[2013-01-31 17:13:19] WARN source/source_read Nothing received on /mount1 for 3 seconds
[2013-01-31 17:13:19] WARN source/source_read Nothing received on /mount1 for 3 seconds
[2013-01-31 17:13:20] WARN source/source_read Nothing received on /mount2 for 3 seconds
[2013-01-31 17:13:20] WARN source/source_read Nothing received on /mount2 for 3 seconds
[2013-01-31 17:13:29] WARN source/source_read Nothing received on /mount3 for 3 seconds
[2013-01-31 17:13:29] WARN source/source_read Nothing received on /mount1 for 3 seconds
[2013-01-31 17:13:29] WARN source/source_read Nothing received on /mount3 for 3 seconds
[2013-01-31 17:13:29] WARN source/source_read Nothing received on /mount1 for 3 seconds
[2013-01-31 17:13:29] WARN source/source_read Nothing received on /mount3 for 3 seconds
[2013-01-31 17:13:30] WARN source/source_read Nothing received on /mount4 for 3 seconds
[2013-01-31 17:13:30] WARN source/source_read Nothing received on /mount4 for 3 seconds
[2013-01-31 17:13:30] WARN source/source_read Nothing received on /mount4 for 3 seconds |
There are approximately 80 mount points on this server, and about half of them were active during this time period. Not every one of the 40 active mount points would throw the 3 second error, but only a small subset of them in a very small amount of time. Each of the four mount points above were sourced from separate geographical areas.
I do not yet have evidence to suggest that there is a problem with the networking equipment of the server this instance is running on, and I don't have an indication yet that it was a networking issue in the immediate upstream vicinity (service provider and datacenter).
What could be causing such a transient issue with small clusters of mount points? |
|
| Back to top |
|
 |
karlH
Code Warrior
Joined: 13 Jun 2005
Posts: 5476
Location: UK
|
| Posted: Sat Feb 02, 2013 12:20 pm Post subject: |
|
|
You can get duplicate references to the same mountpoint at the same time, that is just down to multiple checks during that same second. The fact that not all your mountpoints are affected at the same time will tend to indicate that it is not your end of things. It may be down to a particular route getting congested. I doubt that it will be a busy CPU issue without to noticing other problems on the system.
karl. |
|
| Back to top |
|
|
Nonapeptide
Joined: 08 Sep 2012
Posts: 40
Location: Scottsdale, Arizona
|
| Posted: Fri Feb 08, 2013 2:58 am Post subject: |
|
|
| karlH wrote: |
You can get duplicate references to the same mountpoint at the same time, that is just down to multiple checks during that same second. The fact that not all your mountpoints are affected at the same time will tend to indicate that it is not your end of things. It may be down to a particular route getting congested. I doubt that it will be a busy CPU issue without to noticing other problems on the system.
karl. |
Thanks for the response, Karl!
Good to hear that you suspect it is not a problem with my equipment. The server has a single Xeon E5420 that barely goes above 10%, and when it does it's normally a basic system service like Windows Updates or some kind of maintenance task. No correlation between any kind of CPU "spike" (I hesitate to call it that since the CPU has never been seen to get close to 100%) and the lack of reception on certain mount points has been made yet. Strange transient issue.  |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
