<html><body><br><div dir="ltr">On Aug 27, 2024 3:48 PM, Antoine Beaupré <anarcat@debian.org> wrote:</div>
<div dir="ltr">></div>
<div dir="ltr">> Hey Jerome, </div>
<div dir="ltr">></div>
<div dir="ltr">> I talked with zigo over IRC and will handle this in the git repo, </div>
<div dir="ltr">> reverting back to the 30 days, and addressing the lack of upgrade path. </div>
<br><div dir="ltr">As I wrote on IRC, 30 days is NOT acceptable.</div>
<br><div dir="ltr">On a reasonably large OpenStack cluster, just 2 days of reports means 6GB. Do the math: 30 days would mean nearly 100GB of reports. My puppet VMs, I set then usually with 40GB of system disks...</div>
<br><div dir="ltr">Also, what will you do with 48 puppet report per server per day? Probably nothing...</div>
<br><div dir="ltr">Plaase do NOT revert to 30 days. This is extremely armful for no reason.</div>
<br><div dir="ltr">Thomas Goirand (zigo)</div>
<br><div dir="ltr">P.S: my commit leaves 2 days of reports, not 1, even if it reads "ctime 1"...</div>
<br><br></body></html>