|
| 1 | + |
| 2 | + |
| 3 | + |
| 4 | + |
| 5 | + |
| 6 | +[#breaking-changes-80] |
| 7 | +== Breaking changes 8.0 (August 2025) |
| 8 | + |
| 9 | +Couchbase Server 8.0 was released in August 2025. |
| 10 | +This maintenance release contains the following breaking changes: |
| 11 | + |
| 12 | + |
| 13 | + |
| 14 | + |
| 15 | + |
| 16 | + |
| 17 | + |
| 18 | + |
| 19 | +=== Storage |
| 20 | + |
| 21 | +[#breaking-changes-80-storage] |
| 22 | + |
| 23 | + |
| 24 | +*https://jira.issues.couchbase.com/browse/MB-9418/[MB-9418]*:: |
| 25 | + |
| 26 | +In the latest update, secondary warmup continues paging in items after minimum thresholds are reached, providing more flexibility in configuring warmup behavior. This release includes a new parameter, "warmup_behavior," which allows users to select between blocking, background, or no warmup modes. In "background" mode, value loading occurs concurrently, enabling faster bucket availability. Users can switch between the new default of 0% primary and 100% secondary configuration and the previous behavior, 100% primary and 0% secondary. To simplify configuration while maintaining flexibility, advanced settings remain available through CLI commands for expert users. This update aims to reduce resource waste and improve runtime efficiency, supporting dynamic configurations for various operational needs. |
| 27 | + |
| 28 | +// Generated by [chatgpt:gpt-4o] |
| 29 | + |
| 30 | + |
| 31 | + |
| 32 | +=== Data Service |
| 33 | + |
| 34 | +[#breaking-changes-80-data-service] |
| 35 | + |
| 36 | + |
| 37 | +*https://jira.issues.couchbase.com/browse/MB-63946/[MB-63946]*:: |
| 38 | + |
| 39 | +When using `mcstat` it is no longer possible to use `--json=pretty`. |
| 40 | +The request to pretty-print the JSON is ignored |
| 41 | ++ |
| 42 | +Users can use a pipe in conjunction with an external formatting tool such as `jq` to obtain formatted output: |
| 43 | ++ |
| 44 | +[source, shell] |
| 45 | + |
| 46 | +---- |
| 47 | +
|
| 48 | +./mcstat [options] --json [statgroup] \| jq . |
| 49 | +
|
| 50 | +---- |
| 51 | + |
| 52 | +*https://jira.issues.couchbase.com/browse/MB-63938/[MB-63938]*:: |
| 53 | + |
| 54 | +Build couchbase-server-8.0.0-1964 and couchbase-columnar-1.2.0-1009 now implement a change where mcstat accepts only a single stat group at a time. This update is based on kv_engine commit [0e2fa3c], addressing issue MB-63938. The modification ensures enhanced precision and avoids potential errors in statistical data retrieval. This alignment with the specified commit improves operational consistency across related builds. |
| 55 | + |
| 56 | +// Generated by [chatgpt:gpt-4o] |
| 57 | + |
| 58 | + |
| 59 | +*https://jira.issues.couchbase.com/browse/MB-9418/[MB-9418]*:: |
| 60 | + |
| 61 | +In the latest update, secondary warmup continues paging in items after minimum thresholds are reached, providing more flexibility in configuring warmup behavior. This release includes a new parameter, "warmup_behavior," which allows users to select between blocking, background, or no warmup modes. In "background" mode, value loading occurs concurrently, enabling faster bucket availability. Users can switch between the new default of 0% primary and 100% secondary configuration and the previous behavior, 100% primary and 0% secondary. To simplify configuration while maintaining flexibility, advanced settings remain available through CLI commands for expert users. This update aims to reduce resource waste and improve runtime efficiency, supporting dynamic configurations for various operational needs. |
| 62 | + |
| 63 | +// Generated by [chatgpt:gpt-4o] |
| 64 | + |
| 65 | + |
| 66 | + |
| 67 | +=== Views |
| 68 | + |
| 69 | +[#breaking-changes-80-views] |
| 70 | + |
| 71 | + |
| 72 | +*https://jira.issues.couchbase.com/browse/MB-67280/[MB-67280]*:: |
| 73 | + |
| 74 | +The Bucket Priority option, which previously enabled prioritizing tasks for certain buckets, has been removed from the CB Server UI due to a backend change in the task scheduler since version 7.0. This change was a result of transitioning to FollyExecutorPool, which does not utilize bucket priority in its scheduling algorithm, unlike the earlier CB3ExecutorPool. The Bucket Priority setting is maintained in the REST API to preserve backward compatibility. The old system used to categorize tasks into low and high priority queues, influencing task scheduling when there were more tasks than available threads. With the increased thread pool size in version 7.0, this issue has been alleviated. |
| 75 | + |
| 76 | +// Generated by [chatgpt:gpt-4o] |
| 77 | + |
| 78 | + |
| 79 | +*https://jira.issues.couchbase.com/browse/MB-63212/[MB-63212]*:: |
| 80 | + |
| 81 | +The UI has been updated to remove the option to create memcached buckets. Attempting to create a memcached bucket will now result in no action being taken, although an error should ideally be displayed. While this update is reflected in Morpheus, older versions may not handle the situation correctly without further changes. The UI is designed to prevent errors by checking permissions and inputs before allowing bucket creation. The change is included in the couchbase-server build 8.0.0-1867 and verified in version 8.0.0-2016. This update also appears in couchbase-columnar build 1.2.0-1009 with the relevant commit. |
| 82 | + |
| 83 | +// Generated by [chatgpt:gpt-4o] |
| 84 | + |
| 85 | + |
| 86 | + |
| 87 | + |
| 88 | + |
| 89 | + |
| 90 | + |
| 91 | + |
| 92 | + |
| 93 | + |
| 94 | + |
| 95 | +=== Index Service |
| 96 | + |
| 97 | +[#breaking-changes-80-index-service] |
| 98 | + |
| 99 | + |
| 100 | +*https://jira.issues.couchbase.com/browse/MB-63192/[MB-63192]*:: |
| 101 | + |
| 102 | +The latest build of couchbase-columnar-1.2.0-1009 includes significant changes to shard assignment algorithms for both vector and non-vector indexes, aiming to improve performance by more aggressively sharing plasma shards. This update introduces a configurable feature to revert to previous behavior if needed. During upgrades from versions 7.6 or 7.2 to 8.0, the new index-shard mapping will only be applied upon complete upgrade or during specific rebalance scenarios. The release accommodates Capella by potentially creating additional shards if needed, while ensuring system stability. Further testing is planned to assess performance impact, and ongoing parameter tuning is necessary before finalizing performance tests. |
| 103 | + |
| 104 | +// Generated by [chatgpt:gpt-4o] |
| 105 | + |
| 106 | + |
| 107 | + |
| 108 | + |
| 109 | + |
| 110 | +=== Tools |
| 111 | + |
| 112 | +[#breaking-changes-80-tools] |
| 113 | + |
| 114 | + |
| 115 | +*https://jira.issues.couchbase.com/browse/MB-63389/[MB-63389]*:: |
| 116 | + |
| 117 | +The removal of memcached buckets from couchbase-cli is now implemented. Memcached bucket creation will fail in versions 8.0 and above, with users being advised to use ephemeral buckets instead. To maintain compatibility with older clusters, it is important to incorporate version checks. This change is now reflected in builds couchbase-server-8.0.0-2117 and above, and the couchbase-cli is included in the admin tools package, separate from the server install. The couchbase-cli bucket-create command still lists memcached as an option, but creating such buckets will result in an error message confirming their removal. This update ensures that the couchbase-cli remains functional when managing newer environments. |
| 118 | + |
| 119 | +// Generated by [chatgpt:gpt-4o] |
| 120 | + |
| 121 | + |
| 122 | +*https://jira.issues.couchbase.com/browse/MB-60337/[MB-60337]*:: |
| 123 | + |
| 124 | +In this release, users are prevented from removing backups that would lead to invalid incremental backups. This change is enabled by default to maintain backup integrity. However, users who prefer the previous functionality can utilize a new flag, `--disable-safe-remove-check`, to override this behavior. This update is included in the latest build and addresses the identified issue effectively. Verification was completed on build 8.0.0-2800, and while there is one minor bug, it will be tracked separately. The overall implementation has been validated and deemed satisfactory. |
| 125 | + |
| 126 | +// Generated by [chatgpt:gpt-4o] |
| 127 | + |
| 128 | + |
| 129 | + |
| 130 | +=== Uncategorized |
| 131 | + |
| 132 | +[#breaking-changes-80-uncategorized] |
| 133 | + |
| 134 | + |
| 135 | +*https://jira.issues.couchbase.com/browse/MB-63946/[MB-63946]*:: |
| 136 | + |
| 137 | +When using `mcstat` it is no longer possible to use `--json=pretty`. |
| 138 | +The request to pretty-print the JSON is ignored |
| 139 | ++ |
| 140 | +Users can use a pipe in conjunction with an external formatting tool such as `jq` to obtain formatted output: |
| 141 | ++ |
| 142 | +[source, shell] |
| 143 | + |
| 144 | +---- |
| 145 | +
|
| 146 | +./mcstat [options] --json [statgroup] \| jq . |
| 147 | +
|
| 148 | +---- |
| 149 | + |
| 150 | +*https://jira.issues.couchbase.com/browse/MB-63938/[MB-63938]*:: |
| 151 | + |
| 152 | +Build couchbase-server-8.0.0-1964 and couchbase-columnar-1.2.0-1009 now implement a change where mcstat accepts only a single stat group at a time. This update is based on kv_engine commit [0e2fa3c], addressing issue MB-63938. The modification ensures enhanced precision and avoids potential errors in statistical data retrieval. This alignment with the specified commit improves operational consistency across related builds. |
| 153 | + |
| 154 | +// Generated by [chatgpt:gpt-4o] |
| 155 | + |
| 156 | + |
| 157 | +*https://jira.issues.couchbase.com/browse/MB-9418/[MB-9418]*:: |
| 158 | + |
| 159 | +In the latest update, secondary warmup continues paging in items after minimum thresholds are reached, providing more flexibility in configuring warmup behavior. This release includes a new parameter, "warmup_behavior," which allows users to select between blocking, background, or no warmup modes. In "background" mode, value loading occurs concurrently, enabling faster bucket availability. Users can switch between the new default of 0% primary and 100% secondary configuration and the previous behavior, 100% primary and 0% secondary. To simplify configuration while maintaining flexibility, advanced settings remain available through CLI commands for expert users. This update aims to reduce resource waste and improve runtime efficiency, supporting dynamic configurations for various operational needs. |
| 160 | + |
| 161 | +// Generated by [chatgpt:gpt-4o] |
| 162 | + |
| 163 | + |
0 commit comments