0.7.4. I just got here a few weeks ago and have plans to move everything
up to 1.9.2, which I'm running myself. I'm probably a little premature
asking about what to look for here. I'll try to ask more direct
questions next. Thanks.
On 6/19/19 9:17 AM, Mark Payne wrote:
Russell,
I would also be curious what version of NiFi you are running. Version
1.9.1 introduced a bug [1] that resulted
in the content repository not being properly cleaned up, which could
cause you to run out of disk space.
Thanks
-Mark
[1] https://issues.apache.org/jira/browse/NIFI-6150
On Jun 19, 2019, at 11:12 AM, Russell Bateman <[email protected]
<mailto:[email protected]>> wrote:
Joe,
I am at the beginning of this, I admit. It's useful to understand the
bit about combining content. Our application is for ETL of medical
documents, contains myriad and complex interactions between points,
the documents leave NiFi at points, then return at others before
finally being entrusted to a database/search engine. I'm trying to
corral information on the flow, but it's confusing and hard to know
just where to start my observations, what to look for, etc. This is
why I ask. Thanks.
On 6/19/19 8:58 AM, Joe Witt wrote:
Russell
If data remains in the content repository beyond the specified
archive values then it suggests there is content remaining in the
flow that is not yet eligible to be removed/deleted. This is not
always a direct "500 MB of content waiting for delivery results in
500 MB of content in the content repos" though. That is due to how
the content archiving works and that we tend to combine the content
of many flowfiles into a single physical file on the file system.
It would be necessary to understand a great deal more detail about
your case, flow, config to help more specifically.
Thanks
On Wed, Jun 19, 2019 at 10:54 AM Russell Bateman
<[email protected] <mailto:[email protected]>> wrote:
Just in general, when this data begins to collect without
clearing itself out, what direction might I be looking for the
cause? Ordinarily, in our application, content doesn't collect
flooding the disk and threaten to bring the server down.
Thanks for any comments.