Couchbase server timeout expired after upgrade from 2.5.1 to 3.1.0

[Gleb Chermennov]

We recently upgraded our cluster from Couchbase 2.5.1 to 3.1.0. <br />But when trying to update documents or read from an index, we're getting 'timeout expired' errors. That happens from .NET client and curl/postman. We use .NET client 2.1.2.

<br />

When I looked into the logs, I found errors for document updates in couchdb.log looking like this: `couch_log:error:44]dcp client (bucketName, mapreduce_view: bucketName _design/main (prod/main)): Obtaining mutation from server timed out after 60.0 seconds [RequestId 318, PartId 825, StartSeq 0, EndSeq 34450]. Waiting...` <br />

Something is wrong with compacting/updating indexes: `Set view `bucketName`, main (prod) group `_design/main`, compactor process <0.27790.45> died with unexpected reason: {updater_died,

                                                                                                                          {updater_error,

                                                                                                                           {view_group_index_updater_exit,

                                                                                                                            96,

                                                                                                                            <<>>}}}`

What does it mean? How do I fix it?

[Gleb Chermennov]

Forgot to mention, this bucket is quite big (170 Gb of documents across the cluster) and most of the views are attached to one design document (both dev and production)

вторник, 29 сентября 2015 г., 12:07:57 UTC+3 пользователь Gleb Chermennov написал: