<HTML><BODY><div class="js-helper js-readmsg-msg">
<style type="text/css"></style>
<div>
<base target="_self" href="https://e.mail.ru/">
<div id="style_15723565550071220777_BODY"><div class="class_1572390756">
Thanks!<br><br><p>It seems to be the best solution, solving the problem of ungrounded<br>wal_begin_checkpoint usage and in the same time providing direct vlock<br>obtainment. Sent v3 of the patch.</p><br><br><blockquote style="border-left:1px solid #0857A6;margin:10px;padding:0 0 0 10px;">
Вторник, 29 октября 2019, 0:10 +03:00 от Konstantin Osipov <<a href="mailto:kostja.osipov@gmail.com">kostja.osipov@gmail.com</a>>:<br>
<br>
<div id="">
<div class="js-helper_mailru_css_attribute_postfix js-readmsg-msg_mailru_css_attribute_postfix">
<style></style>
<div>
<div id="style_15722970101087285870_BODY_mailru_css_attribute_postfix">* Vladislav Shpilevoy <<a href="//e.mail.ru/compose/?mailto=mailto%3av.shpilevoy@tarantool.org" target="_blank" rel=" noopener noreferrer">v.shpilevoy@tarantool.org</a>> [19/10/29 00:08]:<br>
> It looks wrong, that we do begin_checkpoint() but omit<br>
> commit_checkpoint(). I would call commit too. Probably now<br>
> begin() does not allocate any resources, but it may start<br>
> in future, and we will have a leak here.<br>
<br>
This would be wrong, too. Wal uses commit_checkpoint() as a singal<br>
that it can recycle wal logs. This is the code Georgy is moving to<br>
GC (but his patch is not in yet).<br>
<br>
One way to fix it would be to add vclock out parameter to<br>
wal_sync().<br>
<br>
-- <br>
Konstantin Osipov, Moscow, Russia<br>
</div>
</div>
</div>
</div>
</blockquote>
<br>
<br>-- <br>Ilya Kosarev<br>
</div></div>
<base target="_self" href="https://e.mail.ru/">
</div>
</div></BODY></HTML>