<feed xmlns='http://www.w3.org/2005/Atom'>
<title>glusterfs.git/tests/basic/nl-cache.t, branch v3.12.5</title>
<subtitle>GlusterFS is a distributed file-system capable of scaling to several petabytes. It aggregates various storage bricks over Infiniband RDMA or TCP/IP interconnect into one large parallel network file system.</subtitle>
<link rel='alternate' type='text/html' href='https://fedorapeople.org/cgit/anoopcs/public_git/glusterfs.git/'/>
<entry>
<title>nl-cache: Fix a possible crash and stale cache</title>
<updated>2017-06-13T05:01:17+00:00</updated>
<author>
<name>Poornima G</name>
<email>pgurusid@redhat.com</email>
</author>
<published>2017-05-26T10:15:57+00:00</published>
<link rel='alternate' type='text/html' href='https://fedorapeople.org/cgit/anoopcs/public_git/glusterfs.git/commit/?id=7674584fa53944a4e982e217798f31a3d1ef313b'/>
<id>7674584fa53944a4e982e217798f31a3d1ef313b</id>
<content type='text'>
Issue1:
Consider the followinf sequence of operations:
   ...
   nlc_ctx = nlc_ctx_get (inode i1)
   ....... -&gt; nlc_clear_cache (i1) gets called as a part of nlc_invalidate
                                   or any other callers
              ...
              GF_FREE (ii nlc_ctx)
   LOCK (nlc_ctx-&gt;lock);  -&gt;  This will result in crash as the ctx
                              got freed in nlc_clear_cache.

Issue2:
   lookup on dir1/file1 result in ENOENT
   add cache to dir1 at time T1
   ....
   CHILD_DOWN at T2
   lookup on dir1/file2 result in ENOENT
   add cache to dir1, but the cache time is still T1
   lookup on dir1/file2 - should have been served from cache
                          but the cache time is T1 &lt; T2, hence
                          cache is considered as invalid.
So, after CHILD_DOWN the right thing would be to clear the cache
and restart caching on that inode.

Solution:
Do not free nlc_ctx in nlc_clear_cache, but only in inode_forget()
The fix for both issue1 and 2 is interleaved hence sending it as
single patch.

Change-Id: I83d8ed36c049a93567c6d7e63d045dc14ccbb397
BUG: 1458539
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/17453
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Issue1:
Consider the followinf sequence of operations:
   ...
   nlc_ctx = nlc_ctx_get (inode i1)
   ....... -&gt; nlc_clear_cache (i1) gets called as a part of nlc_invalidate
                                   or any other callers
              ...
              GF_FREE (ii nlc_ctx)
   LOCK (nlc_ctx-&gt;lock);  -&gt;  This will result in crash as the ctx
                              got freed in nlc_clear_cache.

Issue2:
   lookup on dir1/file1 result in ENOENT
   add cache to dir1 at time T1
   ....
   CHILD_DOWN at T2
   lookup on dir1/file2 result in ENOENT
   add cache to dir1, but the cache time is still T1
   lookup on dir1/file2 - should have been served from cache
                          but the cache time is T1 &lt; T2, hence
                          cache is considered as invalid.
So, after CHILD_DOWN the right thing would be to clear the cache
and restart caching on that inode.

Solution:
Do not free nlc_ctx in nlc_clear_cache, but only in inode_forget()
The fix for both issue1 and 2 is interleaved hence sending it as
single patch.

Change-Id: I83d8ed36c049a93567c6d7e63d045dc14ccbb397
BUG: 1458539
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/17453
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>nl-cache: add group volume set option for ease of use</title>
<updated>2017-06-12T03:52:16+00:00</updated>
<author>
<name>Poornima G</name>
<email>pgurusid@redhat.com</email>
</author>
<published>2017-06-09T06:01:19+00:00</published>
<link rel='alternate' type='text/html' href='https://fedorapeople.org/cgit/anoopcs/public_git/glusterfs.git/commit/?id=38780ff2d0717cd800a49072879a664b04385fd1'/>
<id>38780ff2d0717cd800a49072879a664b04385fd1</id>
<content type='text'>
Change-Id: Id03643a9598da53051a01ca09e1d2a62bc195ab6
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/17495
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Change-Id: Id03643a9598da53051a01ca09e1d2a62bc195ab6
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/17495
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>Implement negative lookup cache</title>
<updated>2017-04-20T04:18:52+00:00</updated>
<author>
<name>Poornima G</name>
<email>pgurusid@redhat.com</email>
</author>
<published>2017-04-12T09:54:14+00:00</published>
<link rel='alternate' type='text/html' href='https://fedorapeople.org/cgit/anoopcs/public_git/glusterfs.git/commit/?id=561766e45a323be3d5c521fdb47f68d71cea776e'/>
<id>561766e45a323be3d5c521fdb47f68d71cea776e</id>
<content type='text'>
Before creating any file negative lookups(1 in Fuse, 4 in SMB etc.)
are sent to verify if the file already exists. By serving these
lookups from the cache when possible, increases the create
performance by multiple folds in SMB access and some percentage
in Fuse/NFS access.

Feature page: https://review.gluster.org/#/c/16436

Updates #82
Change-Id: Ib1c0e7ac7a386f943d84f6398c27f9a03665b2a4
BUG: 1442569
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/16952
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
Tested-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Before creating any file negative lookups(1 in Fuse, 4 in SMB etc.)
are sent to verify if the file already exists. By serving these
lookups from the cache when possible, increases the create
performance by multiple folds in SMB access and some percentage
in Fuse/NFS access.

Feature page: https://review.gluster.org/#/c/16436

Updates #82
Change-Id: Ib1c0e7ac7a386f943d84f6398c27f9a03665b2a4
BUG: 1442569
Signed-off-by: Poornima G &lt;pgurusid@redhat.com&gt;
Reviewed-on: https://review.gluster.org/16952
Reviewed-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
Tested-by: Pranith Kumar Karampuri &lt;pkarampu@redhat.com&gt;
Smoke: Gluster Build System &lt;jenkins@build.gluster.org&gt;
NetBSD-regression: NetBSD Build System &lt;jenkins@build.gluster.org&gt;
CentOS-regression: Gluster Build System &lt;jenkins@build.gluster.org&gt;
</pre>
</div>
</content>
</entry>
</feed>
