Date: Thu, 14 May 2015 12:07:04 -0400
From: Ravi Krishna <s.ravikrishna@aim.com>
To: tgl@sss.pgh.pa.us
Cc: pgsql-admin@postgresql.org
Message-Id: <14d532dde85-2108-310a1@webstg-m03.mail.aol.com>
In-Reply-To: <21483.1431618933@sss.pgh.pa.us>
References: <21483.1431618933@sss.pgh.pa.us>
Subject: Re: WAL and O_DIRECT
MIME-Version: 1.0
Content-Type: multipart/alternative;
 boundary="----=_Part_297018_631856957.1431619624579"
Precedence: bulk
Sender: pgsql-admin-owner@postgresql.org

------=_Part_297018_631856957.1431619624579
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


>> However our DB2 folks are raising a concern that PG WAL writes may not be crash safe, unless we are using
>> write back technology in SAN or SSD , which we are using.

> What's your point exactly?  If the underlying hardware does not provide durable writes, there's
> nothing PG (or DB2) can do to fix that.


Am I right in concluding that PG WAL writes without underlying h/w caching is not crash proof. 
Fortunately these days caching is  ubiquitous in all SSD/SAN technology. Both Oracle and DB2 always open WAL 
logs in O_DIRECT. Is this thinking outdated with modern technology which caches writes. Wonder why Oracle/DB2 
are not making O_DIRECT optional. I am sure it will increase the write performance.


------=_Part_297018_631856957.1431619624579
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit

<font color='black' size='3' face='Verdana, Arial, Helvetica, sans-serif'>
<div style="font-family:arial,helvetica;font-size:10pt;color:black">

<div id="AOLMsgPart_2_d6b6f3b8-fb18-4f41-8c72-ef7bd6a1ae0c">
<font color="black" size="3" face="Verdana, Arial, Helvetica, sans-serif">

<div id="AOLMsgPart_1_2d1a835e-f73a-45e9-8f2d-1a13f8bba508" style="margin: 0px; background-color: rgb(255, 255, 255);"><pre><tt><font face="Tahoma, Verdana, Arial, sans-serif" style="color: rgb(0, 0, 0); font-size: 9pt;">
&gt;&gt; However our DB2 folks are</font><font face="Verdana, Arial, Helvetica, sans-serif" style="color: rgb(0, 0, 0); font-size: 9pt;"> </font><font face="Tahoma, Verdana, Arial, sans-serif"><span style="font-size: 9pt;">raising a concern that PG WAL writes may not be crash safe, unless we are using
&gt;&gt; write back technology in SAN or SSD , which we are using.

&gt; What's your point exactly?  If the underlying hardware does not provide durable writes, there's
&gt; nothing PG (or DB2) can do to fix that.

</span></font></tt><pre style="font-size: 13.3333330154419px;"><tt><tt><font face="Tahoma, Verdana, Arial, sans-serif"><span style="font-size: 9pt;">Am I right in concluding that PG WAL writes without underlying h/w caching is not crash proof. 
Fortunately these days caching is  ubiquitous in all SSD/SAN technology. </span></font></tt><span style="font-size: 9pt; font-family: Tahoma, Verdana, Arial, sans-serif;">Both Oracle and DB2 always open WAL 
logs in O_DIRECT.</span><span style="font-family: Tahoma, Verdana, Arial, sans-serif; font-size: 9pt;"> Is this thinking outdated with modern technology </span><span style="font-size: 9pt; font-family: Tahoma, Verdana, Arial, sans-serif;">which caches writes. Wonder why Oracle/DB2 
are not making O_DIRECT optional. I am sure it will increase the write performance.


</span></tt></pre></pre>
</div>

 
</font>
</div>

</div>
</font>
------=_Part_297018_631856957.1431619624579--