RiakError with no further details sometimes raised when calling store() [JIRA: CLIENTS-553] #410
Comments
Basho-JIRA
changed the title
|
@dellsystem, a try/catch block would certainly be a good solution. You could read the key again and see if it was actually successfully written. It does seem strange, however, that your exception doesn't seem to have any error message. Do you see anything on the server side in the logs? |
|
Thanks for the tip about checking the logs @javajolt. Here's what I see in console.log, at around the same time that the RiakError occurred (different machines with some clock drift, so I'm not sure if it was exactly the same time or not): 2015-08-24 17:05:23.139 [info] <0.419.0>@riak_kv_entropy_manager:perhaps_log_throttle_change:853 Changing AAE throttle from 10 -> 0 msec/key, based on maximum vnode mailbox size 1 from '[email protected]' Would that explain it? There are no other logs that contain entries around that time. |
|
I don't think that would be problematic. It just information from the AAE subsystem. |
|
Could it have something to do with Solr? The bucket that I was updating has a Solr index, and I've been having some problems with Solr lately - it seems as if there's a Riak node containing old versions of certain search indexes. I also see this in /var/log/solr.log of some (but not all) of the Riak nodes in my cluster: This message shows up on the logs approximately once a minute. Not sure what's causing it, or if it's even related to my original problem, so any thoughts on this would be appreciated |
|
Are you re-creating the Solr index? If so, we already have an issue pointing out this problem: #408 |
|
Not that I know of. I do have scripts that occasionally try to access the indexes mentioned, but they never deliberately attempt to create the index, and they don't run at that frequency either. Is there anything else that could be causing those errors? I've also recently been getting a RiakError with no further details when trying to run a Solr query using the Riak python client, but this time I see the following message in solr.log. Maybe it's related? EDIT: Just had a thought, related to the issue you linked. I did do some index deleting/recreating a while back. Could the log messages be related to that? Is there a background process that periodically checks search indexes for validity or something, that keeps running even weeks after the indexes were last updated? |
|
@dellsystem Which version of Riak are you using? Could you explicitly remove this index (someindexname) especially if you are not using it? |
|
@javajolt, 2.1.1. I'm going to do some index deleting/recreating to see if that fixes anything. I'll update this issue with what I find. Thanks for your help. |
|
Update on this, in case anyone else has the same problem: I've rebuilt all the problematic indexes, upgraded the specs of the machines in the cluster, and incorporated the retrying Python library in case it happens again. I haven't seen this specific error (when calling |
|
@dellsystem Can we please close this issue then and reopen if it comes back? I'm also suspicious that this might be related to Solr than to the Python client. |
|
It does feel like more of a Solr issue. Will do. |
|
Fixed, or closed via GitHub issues. [posted via JIRA by Alexander Moore] |
Very occasionally, I'll see a RiakError (with no further details) raised during the execution of my code, typically when calling the
store()method on a riak object (with no arguments).The traceback (starting from the line immediately after the
store()call) looks like this:It happens very rarely, and I haven't been able to reproduce it on demand yet to uncover the actual problem. In the meantime, I would appreciate any suggestions on the best way to handle this. Should I catch the RiakError in the application level, wrapping the
.store()call in a try/catch block? Or is there a better way?Thanks in advance.
The text was updated successfully, but these errors were encountered: