Hi Rob, I tried both suggestions and worked as expected. These two examples help me have much better understanding of the data flow in heka.
Thanks a lot. Ming -----Original Message----- From: Rob Miller [mailto:[email protected]] Sent: Wednesday, August 26, 2015 4:45 PM To: Ho, Ming; [email protected] Subject: Re: [heka] Not getting LogOutput from linux_memstats.lua Your setup is working exactly as it should. When the data is run through the MemStatsDecoder, it is parsed from the message payload and stored in various message fields. By default the payload is then removed from the message. Your LogOutput is using a PayloadEncoder, so it is dutifully showing the empty payload field. You'll be able to see the original payload value if you add a `[MemStatsDecoder.config]` section and set `payload_keep = true`. Even better, though, would be to output the whole message so you can see all of the fields and not just the payload. To do this you can switch from using a PayloadEncoder to using the RstEncoder (http://hekad.readthedocs.org/en/v0.10.0b1/config/encoders/rst.html). -r On 08/26/2015 11:50 AM, Ho, Ming wrote: > Hi Rob, > I have a project to take a look at heka. I am new to heka. I got it > installed and running a couple days ago. I am able to use heka to do > some very simple log watching. Now I try (just to learn as an example) > the following syntax in my config.toml and expect to see some data in > stdout. > [MemStats] > type = "FilePollingInput" > ticker_interval = 5 > file_path = "/proc/meminfo" > decoder = "MemStatsDecoder" > [MemStatsDecoder] > type = "SandboxDecoder" > filename = "lua_decoders/linux_memstats.lua" > [PayloadEncoder] > #append_newlines = false > append_newlines = true > # LogOutput plugin ouput to stdout > [LogOutput] > message_matcher = "TRUE" > encoder = "PayloadEncoder" > But when I run hakad with this, all I get is the date time every 5 > seconds without any data/,/ e.g.: > 2015/08/26 18:37:04 > 2015/08/26 18:37:09 > 2015/08/26 18:37:14 > 2015/08/26 18:37:19 > What am I doing wrong? My final goal is to use "Apache Access Log > Decoder" to capture access.log and then use "ElasticSearch JSON Encoder" > to send data to ElasticSearch (I know it is a long way to go for me!) > Appreciate your help in advance. > Thanks. > Ming Ho _______________________________________________ Heka mailing list [email protected] https://mail.mozilla.org/listinfo/heka
