XDCR: UI should show the reason why replication failed

Junyi Xie added a comment - 19/Jun/12 4:01 PM Yes, we dump logs of the error raised from Couchdb replicator and XDCR manager on why the replication failed. I think we just need to expose that to UI. Filed a bug on that. Thanks, Junyi On Jun 19, 2012, at 1:50 PM, Sharon Barr wrote: Can you first make sure there is a reason somewhere in the logs. And then open a bug on this.   Sharon   From: Junyi Xie [mailto: junyi.couchbase@gmail.com ] On Behalf Of Junyi Xie Sent: Tuesday, June 19, 2012 1:47 PM To: Sharon Barr Cc: Ketaki Gangal; Farshid Ghods; Aliaksey Kandratsenka; Chiyoung Seo; Damien Katz Subject: Re: Replication status is failed on the UI   In this case probably a couple of lines on UI describing the reason of failure would be helpful. We need to expose the reason of failure from XDCR to UI   Junyi   On Jun 19, 2012, at 1:42 PM, Sharon Barr wrote: I am talking about the fact that the UI show fail status and the user have no clue what failed. User visable logs also shows nothing interesting (10.2.1.12). Sharon On Jun 19, 2012, at 1:15 PM, "Junyi Xie" < junyi@couchbase.com > wrote: Which bug are you referring to? XDCR could fail due to different reasons, e.g., unable to reach remote cluster, vbucket map change, etc.   Need to ask Alk what status from XDCR rep manager will make UI show "failed".   CC Damien.   Thanks,   Junyi     On Jun 19, 2012, at 9:31 AM, Sharon Barr wrote: Junyi,   There is no point of just opening a bug on this. What logs do you need to understand why the status is failed, and what does it mean??   <image001.png>    

Peter Wansch added a comment - 15/Aug/12 1:34 PM Damien's rework will provide the error messages.

Junyi Xie added a comment - 05/Sep/12 2:34 PM Aaron, Since Damien and I are busy working remove blockers, could you please help us work on this? This is not a blocker but we hope fix it in 2.0. Thanks!

Thuan Nguyen added a comment - 10/Sep/12 4:24 PM Integrated in github-ns-server-2-0 #464 (See [ http://qa.hq.northscale.net/job/github-ns-server-2-0/464/ ])      MB-5611 : Record last 10 errors per xdc replication (Revision e5d3a4ced105dcc2ebf9d38d14df6de4d3af6f8f)      Result = SUCCESS pwansch : Files : * src/xdc_replication_sup.erl * src/xdc_vbucket_rep.erl * src/misc.erl * src/xdc_rep_manager.erl * src/xdc_rep_utils.erl * src/xdc_replication.erl

Thuan Nguyen added a comment - 10/Sep/12 4:24 PM Integrated in github-couchdb-preview #504 (See [ http://qa.hq.northscale.net/job/github-couchdb-preview/504/ ])      MB-5611 : Reduce replication http retries from 10 to 2 (Revision a672a1ddb9c41cbf7132ca864d24e0dd54889be7)      Result = SUCCESS pwansch : Files : * etc/couchdb/default.ini.tpl.in

Aaron Miller added a comment - 10/Sep/12 4:39 PM Also working on blockers, I can't put time on this if it's not a blocker.

Junyi Xie added a comment - 13/Sep/12 5:34 PM Pavel, Since Aaron is busy with other blockers, could you please take a look at this bug since you are working on another similar one ( MB-6275 )? Please let me or Peter know if you have any questions. Thanks!

Aleksey Kondratenko added a comment - 20/Sep/12 4:59 AM Dipti, we haven't planned this for current sprint. Doesn't feel like big task, but it's not trivial either

Ketaki Gangal added a comment - 26/Sep/12 5:10 PM Add more information on what type of errors to display on the UI.

Abhinav Dangeti added a comment - 28/Sep/12 3:40 PM Some common XDCR failure reasons: 1 db_not_found error: when node is unresponsive, for e.g: "could not open http://Administrator: *****@10.3.3.28:8092/default%2f120%3b093c0a978eb59342ea52d87eae424bb3/" 2 badmatch,{error,corrupted_data}, Erlang-related corruption     [{couch_compress,decompress,1},      {couch_doc,with_uncompressed_body,1},      {couch_doc,to_json_base64,1},      {xdc_vbucket_rep_worker,maybe_flush_docs,3},      {lists,foldl,3},      {xdc_vbucket_rep_worker,local_process_batch,5},      {xdc_vbucket_rep_worker,queue_fetch_loop,4}] 3 checkpoint_commit_failure      {bad_return_value,        {checkpoint_commit_failure,            <<"Failure on target commit: {error,<<\"not_found\">>}">>}} 4 http_request_failed xdc_replicator:handle_info:282] Worker <0.11173.72> died with reason: {http_request_failed,"POST",                                        " http://10.3.121.33:8092/default%2F684/_bulk_docs ",                                        {error,{code,500}}}   Replicator: couldn't write document xdc_replicator_worker:flush_docs:111] Replicator: couldn't write document ``, revision ``, to target database ` http://10.3.121.33:8092/default%2F683/ `. Error: ``, reason: ``. 5 replicator_died {replicator_died, {'EXIT',<15849.2212.0>, {badmatch,{error,closed}}}} 6 bulk_set_vbucket_state_failed General error seen when rebalance fails due to vbucket_map not ready (possibly) that may cause replication to fail.

Junyi Xie added a comment - 02/Oct/12 4:58 PM We already have a few bugs filed for XDCR error/warning msg on UI, Let's put everything in one place holder ( MB-6763 ).

Junyi Xie added a comment - 02/Oct/12 4:58 PM MB-6763