Getting Data In

post 4.3 upgrade - unable to read raw size

Contributor

We just upgraded our environment to 4.3.2 from 4.2.3 and see this message spamming our splunkd.logs on both hot/warm and cold data partitions

05-08-2012 00:33:30.407 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data/splunk/var/lib/splunk/transient/db/db_1335815136_1335613695_110/.rawSize": No such file or directory
05-08-2012 00:33:30.407 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data/splunk/var/lib/splunk/transient/db/db_1335957976_1335815156_111/.rawSize": No such file or directory
05-08-2012 00:33:30.407 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data/splunk/var/lib/splunk/transient/db/db_1336089130_1335957977_112/.rawSize": No such file or directory

05-08-2012 00:33:58.700 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data2/colddb/os/db_1323061363_1323023831_1192/.rawSize": No such file or directory
05-08-2012 00:33:58.714 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data2/colddb/os/db_1322956802_1317593989_1193/.rawSize": No such file or directory
05-08-2012 00:33:58.767 +0000 INFO  timeinvertedIndex - Unable to read raw size file="/data2/colddb/os/db_1313294562_1311652962_1195/.rawSize": No such file or directory

What does this mean and is there a way to resolve it?

Tags (2)
0 Karma
1 Solution

Splunk Employee
Splunk Employee

[EDIT]

Irrelevant answer, see comment below :

This is a know bug, the message is not an error, but is annoying.
Something is wrong in your bucket manifest. The best solution is to force splunk to recreate it.

for the buckets with the issue :

  • in the index delete the hidden file .bucketManifest
  • in the index/bucket, delete the hidden files .rawSize and .sizeManifest4.1
  • restart splunk, to refresh and regenerate them (may take some minutes)
  • check in the logs if the message comes back.

View solution in original post

Splunk Employee
Splunk Employee

[EDIT]

Irrelevant answer, see comment below :

This is a know bug, the message is not an error, but is annoying.
Something is wrong in your bucket manifest. The best solution is to force splunk to recreate it.

for the buckets with the issue :

  • in the index delete the hidden file .bucketManifest
  • in the index/bucket, delete the hidden files .rawSize and .sizeManifest4.1
  • restart splunk, to refresh and regenerate them (may take some minutes)
  • check in the logs if the message comes back.

View solution in original post

Splunk Employee
Splunk Employee

message fixed in 4.3.4

0 Karma

Splunk Employee
Splunk Employee

Hi SonicZ

It seems that my answer was incorrect. The issue is a known bug SPL-52868, that can occur after an upgrade to 4.3
Please disregard the INFO message, it's not relevant.
It will be fixed in next release.

0 Karma

Path Finder

Didn't help.

0 Karma

Contributor

Good to know, thanks!

0 Karma