Host name extraction via regex on indexing - Only indexing a single file?

[rturk]

Greetings fellow Splunkers,

I'm having some issues with extracting the correct host name from log file names on indexing.

I keep my log files in the following directory:

/var/splunk/input/mms_logs/

The filename structure is:

mms_HOST-IP-ADDRESS_TIMESTAMP.log

examples:

mms_10.152.58.100_20110101_004000_06137.log
mms_10.152.58.194_20110121_120000_70656.log

All log files have identical file permissions: (rw-r--r--).

Now to extract the IP address portion of filename as a host, I used the following regex:

/var/splunk/input/mms_logs/mms_(\d+.\d+.\d+.\d+)_\d+

Now while that works... it only seems to extract the hostname & event data from a single file. I have a single source, sourcetype, and host.

This is despite 10,000+ files being in the directory, and the list of data inputs showing it has detected 12055 files in the directory :confused face:

If someone can shed some light on this it would be greatly appreciated 🙂

[rturk]

Hey meno 🙂 Sorry for not getting back to you earlier, but I didn't notice the small text below my question. I managed to find the answer elsewhere and I have some information to get further clarification.

Thanks 🙂

[rturk · Accepted solution]

Right... a bit of digging around has turned up the goods 🙂

Checked in /var/splunk/logs/splunk/splunkd.log which had multiple instances of the following:

02-23-2011 11:51:25.673 ERROR TailingProcessor - Ignoring path due to: File will not be read, seekptr checksum did not match (file=/var/splunk/input/mms_logs/mms_10.152.58.196_20110206_211500_28809).  Last time we saw this initcrc, filename was different.  You may wish to use a CRC salt on this source.  Consult the documentation or file a support case online at http://www.splunk.com/page/submit_issue for more info.

Now one thing I neglected to mention (and had no idea would be an issue) is that this data had previously been indexed, however I cleaned out the index with:

/opt/splunk/bin/splunk clean eventdata -index main

After searching for "ERROR TailingProcessor - Ignoring path due to: File will not be read, seekptr checksum did not match", I was directed to this question on answers.splunk.com:

http://answers.splunk.com/questions/1568/windows-dhcp-log-files-too-small-to-match-seekptr-checksum

Adding crcSalt = <SOURCE> to the bottom of my inputs.conf and restarting Splunk solved the issue. Now to get my head around why this was needed...

[meno]

I would like to see your inputs.conf, props.conf, transforms.conf, if possible 😉