MongoDB log: extent 0:550000 and can't find plugin [desc] - mongodb

Question

I am using Meteor with a separate MongoDB running on Windows. I ran MongoDB as a service.
My MongoDB log is full of the following:
Mon Mar 04 14:15:36 [conn19] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:15:38 [conn17] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:15:40 [conn16] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:15:43 [conn18] warning: can't find plugin [desc]
Mon Mar 04 14:15:43 [conn19] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:15:49 [conn18] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:16:14 [conn16] warning: can't find plugin [desc]
Mon Mar 04 14:16:20 [conn17] info DFM::findAll(): extent 0:55000 was empty, skipping ahead. ns:webfm.graphdata
Mon Mar 04 14:16:24 [conn16] warning: can't find plugin [desc]
Mon Mar 04 14:16:32 [conn20] info DFM::findAll(): extent 0:60000 was empty, skipping ahead. ns:webfm.history
Mon Mar 04 14:16:34 [conn16] warning: can't find plugin [desc]
From what I can find the "findAll(): extent 0:55000..." seemed to have to do with my application's removing the data frequently. Is that correct?
How about the can't find plugin [desc]? What's that plugin? How can I fix it?

Answer 1

From what I can find the "findAll(): extent 0:55000..." seemed to have to do with my application's removing the data frequently. Is that correct?
Per Nick's answer, the message extent 0:55000 was empty, skipping ahead is related to your frequent deletions. This is a warning emitted when skipping an empty data extent, with 0:55000 being the extent location. The space will normally be reused as you add more data to that collection.
If you are frequently adding and deleting this collection and disk space is of concern, you could also consider:
enabling usePowerof2Sizes flag for this collection (MongoDB 2.2 or newer) for more effective reuse of deleted space
saving temporary collections in a separate database, and dropping the database when done (rather than running frequent compacts or repairs)
How about the can't find plugin [desc]? What's that plugin? How can I fix it?
This typically means you have an invalid index definition and the index "plugin" can't be found to handle that. In this example, I expect you have an indexed field declared with a property of desc rather than -1.
You can check the index definitions with db.collection.getIndexes():
use webfm
db.graphdata.getIndexes()

Answer 2

The message is in fact informational and harmless. It tells you that a whole extent was empty. This can happen if you deleted a lot of data recently.
You can use the compact command to defragment and compact a collection (basically re-writing it to disk and re-creating the indexes on the collection), which would get rid of the log message. Please be aware that compact is a resource-heavy operation.

Answer 3

I'm not absolute sure this is a problem. You might be running mongodb with too much verbosity, so its showing excessive detail in its warnings and processes for debug reasons. This is a good thing, because if theres a problem or bug the issue can be identified.
But if you want to remove it:
Decrease the number of 'v's or remove them in your startup command.
If you use a config file change whichever of these you have to false:
vvvvv = false
vvvv = false
vvv = false
vv = false
v = false