You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello ,
I have a question , I am working on upgrade secor version to be the secor master branch ..
with that I succeed to run "Dev" env ,where I write all output file .gzip to s3 "Dev" bucket ..
what I saw that secor create lot of small files ..in s3 bucket ..
question - how can I tune secor to create bigger size files and as a result of this will have less files number for the same events fetch from kafka ..
for example if reading 1000 events from topic .. , it can be split to 100 files of 10 events each file , or it can be split to 50 files where each file contains 20 events ..
I saw there is "secor.max.file.age.seconds" and "secor.max.file.size.bytes" properties .. but as far as I understood those impact the frequency ("when") the uplaod will be called .. but not the crated file size on disk ..
I am not sure I fully understand how secor "know" when to stop write the current file and start create the new file ..
is it upon file size , kafka batch .. or other crietira ? , file rolling policy ?
how can I control the size of a "single" output log file
Thanks in advance !!
Nir
The text was updated successfully, but these errors were encountered:
Hello ,
I have a question , I am working on upgrade secor version to be the secor master branch ..
with that I succeed to run "Dev" env ,where I write all output file .gzip to s3 "Dev" bucket ..
what I saw that secor create lot of small files ..in s3 bucket ..
question - how can I tune secor to create bigger size files and as a result of this will have less files number for the same events fetch from kafka ..
for example if reading 1000 events from topic .. , it can be split to 100 files of 10 events each file , or it can be split to 50 files where each file contains 20 events ..
I saw there is "secor.max.file.age.seconds" and "secor.max.file.size.bytes" properties .. but as far as I understood those impact the frequency ("when") the uplaod will be called .. but not the crated file size on disk ..
I am not sure I fully understand how secor "know" when to stop write the current file and start create the new file ..
is it upon file size , kafka batch .. or other crietira ? , file rolling policy ?
how can I control the size of a "single" output log file
Thanks in advance !!
Nir
The text was updated successfully, but these errors were encountered: