All of lore.kernel.org
 help / color / mirror / Atom feed
From: YuFan Chen <wiz.chen@gmail.com>
To: ceph-devel@vger.kernel.org
Subject: About the "temperature histogram" of tier-agent
Date: Tue, 28 Nov 2017 14:38:03 +0800	[thread overview]
Message-ID: <CAHMqkF7XOhzrQJODwOGnz-0Y8echtdQPEamcNL126w5ctW4+Xg@mail.gmail.com> (raw)

Hi,

   Before Jewel, the code uses "atime" to distinguish hot/cold data:

          agent_estimate_atime_temp() calculates the distance.
          It means colder data gets longer distance (larger value)


   In Jewel, it changes to use temperature with weight.
            So, colder data gets smaller value.

    However, it still use the "old semantics" to determine evicting or not in
           ReplicatedPG::agent_estimate_atime_temp();

    if (1000000 - temp_upper >= agent_state->evict_effort)
       return false;


     if there is no misunderstanding, should it change to that:

--- a/src/osd/ReplicatedPG.cc
+++ b/src/osd/ReplicatedPG.cc
@@ -12211,7 +12211,7 @@ bool
ReplicatedPG::agent_maybe_evict(ObjectContextRef& obc, bool
after_flush)
     delete f;
     *_dout << dendl;

-    if (1000000 - temp_upper >= agent_state->evict_effort)
+    if (temp_upper >= agent_state->evict_effort)
       return false;
   }

Yu Fan

             reply	other threads:[~2017-11-28  6:38 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-11-28  6:38 YuFan Chen [this message]
2017-11-28 13:20 ` About the "temperature histogram" of tier-agent Sage Weil
2017-11-30  8:51   ` Li Wang
     [not found]     ` <CAHMqkF5VckeHoMy2qhUR5M=ivzujGTaiAoutjW8LMjL7vQSxtQ@mail.gmail.com>
2017-12-01 12:56       ` Li Wang
2017-12-01 15:12         ` YuFan Chen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAHMqkF7XOhzrQJODwOGnz-0Y8echtdQPEamcNL126w5ctW4+Xg@mail.gmail.com \
    --to=wiz.chen@gmail.com \
    --cc=ceph-devel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.