Can you clarify: are you referring to link walking or key retrievals?
In either case, if you have a bucket and key that you believe is tied to bad
data on one of your nodes, you could do something like the following.
First, run riak-admin to generate a full set of cluster details:
$ riak-admin cluster-info /tmp/my-cluster.txt
Then, sift through the resulting data set for all of the HTTP endpoints. The
following one-liner might work, depending on the vagaries of regex support in
your tools.
Replace <bucket> and <key> with the relevant data. The output, if this works,
is a shell script which you can run to query each vnode for that bucket/key
combination.
$ egrep '^\s*{http,' /tmp/my-cluster.txt | sed -E 's/ +//g' | sort | uniq | sed
's/.*{//' | sed 's/}.*//' | sed 's/"//g' | awk 'BEGIN { FS="," } { print "curl
-v http://"; $1 ":" $2 "/riak/<bucket>/<key>" }' > /tmp/runme.sh
The goal is to pick out all occurrences of "{http," from the cluster
information file and run:
curl -v http://<ip>:<port>/riak/<bucket>/<key>
-John Daily
Technical Evangelist
Basho
On Jan 31, 2013, at 12:16 PM, Martin Streicher
<[email protected]> wrote:
> How can I find the queries that Ripple generates? I need to track down an
> issue in production -- at least one machine is producing bad results. I need
> to be able to run the query(ies) that Ripple generates directly on each node
> to narrow.
>
> Martin
>
>
> _______________________________________________
> riak-users mailing list
> [email protected]
> http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
_______________________________________________
riak-users mailing list
[email protected]
http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
