https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 https://jira.whamcloud.com/browse/LU-2925 Lustre <p>this patch (<a href="http://review.whamcloud.com/#change,4940" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/#change,4940</a>) assumes that most MDS services only need 5K as request buffer size, the only exception is "regular" service which can get large request with LOV EA information. <br/> However, OUT service of DNE can have request size up to 9K, which means we also need to define individual buffer size for OUT. </p> LU-2925

OUT service of MDS needs bigger request buffer

Bug Blocker Resolved Fixed Liang Zhen Liang Zhen HB Thu, 7 Mar 2013 03:02:30 +0000 Fri, 15 Mar 2013 08:52:36 +0000 Fri, 15 Mar 2013 08:52:36 +0000 Lustre 2.4.0 Lustre 2.4.0 0 3 <p>wangdi, could you please comment how we calculate OUT request/reply buffer size? So I can put it into comment of patch, thanks</p> <p>Here is update request size</p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>static const struct req_msg_field *mds_update_client[] = {        &amp;RMF_PTLRPC_BODY,        &amp;RMF_UPDATE, }; </pre> </div></div> <p>and RMF_UPDATE is a EA data, with 8K size. </p> <div class="preformatted panel" style="border-width: 1px;"><div class="preformattedContent panelContent"> <pre>#define UPDATE_BUFFER_SIZE 8192 </pre> </div></div> <p>The update req includes all of updates from the create, which might include linkea (4K maxim), together with other updates, we set it to 8K now.</p> <p>patch is here: <a href="http://review.whamcloud.com/#change,5653" class="external-link" target="_blank" rel="nofollow noopener">http://review.whamcloud.com/#change,5653</a></p> <p>Landed for 2.4</p> Development Rank 1|hzvkcv: Rank (Obsolete) 7026 Severity