incubator-blur-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Aaron McCurry <amccu...@gmail.com>
Subject Re: [11/12] git commit: allow blur to start if no commands file found
Date Wed, 29 Oct 2014 11:48:15 GMT
I couldn't get the shard server to not start with the command path not
being set.  I am testing on a mac.  I have added some logging and for the
default directory ($BLUR_HOME/commands) to be created if it's missing and
if it can't be created then it should run with auto pickup of new commands
disabled.

Aaron

On Wed, Oct 29, 2014 at 7:17 AM, Tim Williams <williamstw@gmail.com> wrote:

> On Wed, Oct 29, 2014 at 12:10 AM,  <aavenoso@apache.org> wrote:
> > allow blur to start if no commands file found
> >
> >
> > Project: http://git-wip-us.apache.org/repos/asf/incubator-blur/repo
> > Commit:
> http://git-wip-us.apache.org/repos/asf/incubator-blur/commit/ef79c773
> > Tree:
> http://git-wip-us.apache.org/repos/asf/incubator-blur/tree/ef79c773
> > Diff:
> http://git-wip-us.apache.org/repos/asf/incubator-blur/diff/ef79c773
> >
> > Branch: refs/heads/feature_update_jetty
> > Commit: ef79c773e2000661d20461f3959154bcb9a6c538
> > Parents: 414575e
> > Author: Andrew Avenoso <andrew.va@gmail.com>
> > Authored: Wed Oct 29 00:07:01 2014 -0400
> > Committer: Andrew Avenoso <andrew.va@gmail.com>
> > Committed: Wed Oct 29 00:09:37 2014 -0400
> >
> > ----------------------------------------------------------------------
> >  .../apache/blur/command/BaseCommandManager.java | 20
> +++++++++++---------
> >  1 file changed, 11 insertions(+), 9 deletions(-)
> > ----------------------------------------------------------------------
> >
> >
> >
> http://git-wip-us.apache.org/repos/asf/incubator-blur/blob/ef79c773/blur-core/src/main/java/org/apache/blur/command/BaseCommandManager.java
> > ----------------------------------------------------------------------
> > diff --git
> a/blur-core/src/main/java/org/apache/blur/command/BaseCommandManager.java
> b/blur-core/src/main/java/org/apache/blur/command/BaseCommandManager.java
> > index ba6d4be..f6dfb9d 100644
> > ---
> a/blur-core/src/main/java/org/apache/blur/command/BaseCommandManager.java
> > +++
> b/blur-core/src/main/java/org/apache/blur/command/BaseCommandManager.java
> > @@ -167,16 +167,18 @@ public abstract class BaseCommandManager
> implements Closeable {
> >    protected synchronized int loadNewCommandsFromCommandPath() throws
> IOException {
> >      Path path = new Path(_commandPath);
> >      FileSystem fileSystem = path.getFileSystem(_configuration);
> > -    FileStatus[] listStatus = fileSystem.listStatus(path);
> >      int changeCount = 0;
> > -    for (FileStatus fileStatus : listStatus) {
> > -      BigInteger contentsCheck = checkContents(fileStatus, fileSystem);
> > -      Path entryPath = fileStatus.getPath();
> > -      BigInteger currentValue = _commandPathLastChange.get(entryPath);
> > -      if (!contentsCheck.equals(currentValue)) {
> > -        changeCount++;
> > -        loadNewCommand(fileSystem, fileStatus, contentsCheck);
> > -        _commandPathLastChange.put(entryPath, contentsCheck);
> > +    if(fileSystem.exists(path)) {
> > +      FileStatus[] listStatus = fileSystem.listStatus(path);
> > +      for (FileStatus fileStatus : listStatus) {
> > +        BigInteger contentsCheck = checkContents(fileStatus,
> fileSystem);
> > +        Path entryPath = fileStatus.getPath();
> > +        BigInteger currentValue = _commandPathLastChange.get(entryPath);
> > +        if (!contentsCheck.equals(currentValue)) {
> > +          changeCount++;
> > +          loadNewCommand(fileSystem, fileStatus, contentsCheck);
> > +          _commandPathLastChange.put(entryPath, contentsCheck);
> > +        }
>
> I agree with the intent here but I'm wondering if it should be
> addressed further up so we don't waste a timer task checking it all
> the time?  Also, if it got this far, it means they *attempted* to set
> it so we should probably give at least a WARN about the path?
>
> Thanks,
> --tim
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message