There are many junk tables such as work tables, log tables and backup tables... I want to reject them from the table list.

TableMemo also displays the update interval and last modified date of the table you are looking at.

Now, meta data of table is the column structure and the number of rows, only two types.
Add a statistic information to show how much new rows are in the table.

Hi, we use dmemo.
Thanks a lot!!

By the way, are you planning to update to Rails 5?

Currently I am not actively maintaining it, while @ragi256 is very actively maintaining it.

I think I'm going to transfer it to organization, because it's difficult to maintain in a personal repository because of its limited permissions.

@ragi256 What do you think?

以下の画像のように、サンプルデータのところでtimestamp without time zone型の値に対してUTCと表示されてしまいます。しかも実はこの時刻はJSTなので、すごく紛らわしいという……。

元DBはRedshiftです。

DataSourceとの同期がとにかく遅い

Binary column type (like varbinary) value can not format.

But I have no better idea.
Unpacking?
I just want to see value type or existence when binary column type.
How about my idea ?

I registered Keyword contain comma(ex:Developers.IO).
but this description page is not move.(status of 404).
I use dmemo version 0.3.2.

キーワードに.(ドット)を含むようなキーワードを登録するとDBに登録はされているのですが、
詳細ページに移動しようとするとDBへのSelect文が.(ドット)の手前まででDBに問い合わせているため、レコードが見つからずページ遷移が失敗しています。

登録が行われた際のログと、その後のページ遷移を行おうとした際のログを添付します。

登録時のログ

web_1  |   SQL (0.6ms)  INSERT INTO "keywords" ("name", "description", "created_at", "updated_at") VALUES ($1, $2, $3, $4) RETURNING "id"  [["name", "Developers.IO"], ["description", "省略"], ["created_at", "2016-09-15 03:00:27.302907"], ["updated_at", "2016-09-15 03:00:27.302907"]]
web_1  |   SQL (0.6ms)  INSERT INTO "keyword_logs" ("revision", "user_id", "description", "description_diff", "keyword_id", "created_at") VALUES ($1, $2, $3, $4, $5, $6) RETURNING "id"  [["revision", 1], ["user_id", 2], ["description", "省略"], ["keyword_id", 1], ["created_at", "2016-09-15 03:00:27.306320"]]
web_1  |    (4.8ms)  COMMIT

遷移時のログ

web_1  | Started GET "/keywords/Developers.IO" for 172.19.0.1 at 2016-09-15 03:00:27 +0000
web_1  | Processing by KeywordsController#show as
web_1  |   Parameters: {"id"=>"Developers"}
web_1  |   User Load (0.2ms)  SELECT  "users".* FROM "users" WHERE "users"."id" = $1 LIMIT 1  [["id", 2]]
web_1  |   Keyword Load (0.5ms)  SELECT  "keywords".* FROM "keywords" WHERE "keywords"."name" = $1 LIMIT 1  [["name", "Developers"]]
web_1  | Completed 404 Not Found in 7ms (ActiveRecord: 0.7ms)
web_1  |
web_1  | ActiveRecord::RecordNotFound (Couldn't find Keyword):
web_1  |   app/controllers/keywords_controller.rb:11:in `show'

database_memos -> schema_memos -> table_memos -> column_memos

• Link to re:dash
• Link to esa
• ...

• CI
• [ ] Docker build && push
  • https://docs.docker.com/docker-hub/builds/

I moved Keyword contents ,but Database column is active in sidemenu.
I use dmemo version 0.3.2.

左側のサイドメニューの項目が、
他のページ(Keyword等)に遷移しても必ずDatabaseがアクティブになっています。
Databaseが登録されたない状態のみかとも思ったのですが、
Databaseを登録しても同様の現象です。

Use #with_search_path https://github.com/hogelog/dmemo/blob/master/app/models/data_source.rb#L19
at the suitable positions. (hook?)

https://coveralls.io/github/hogelog/dmemo

Related to #292

There are many other systems around Redshift as a DWH DB.
For example, data pipeline jobs, streaming load, ETL batch.
It would be useful to be able to register various metadata of a table to Dmemo from other systems.

(And if possible, I want to add API to provide metadata.)

If this feature is added, Dmemo can also be a convenient metadata repository, and there will be more ways to use it.

Want to write description to Datasources on Tableau Server.

Add API functionality to Dmemo to make it easier to combine DWH environments.
If Dmemo can be treated not only as a document site but also as an API server, it will be more useful.

For example

• auto complete of table columns for SQL Editor (such as bdash)
• auto link to Dmemo in SQL share web app ( such as bdash server)
• cache layer for meta data of DWH DB

ridgepole:apply in https://github.com/hogelog/dmemo/blob/master/bin/docker_start.sh

やっちゃっていいような気もする。

https://github.com/hogelog/dmemo/blob/master/app/controllers/table_memos_controller.rb#L15

一度登録したキーワードの削除は機能としてはないのでしょうか？

Current DataSource::DynamicTable::* models log is too quiet.

Table names in markdown document should become hyperlink to the document page.

Hello,

First of all, Thank you for supporting Presto adopter. I'm happy it because I gave up it when dmemo was released.
I have a question, so I file this issue. If you have a chance, please give me your advice.

Description

I tried our Presto environment (TreasureData), which requires SSL parameter with 443 port, to connect from dmemo.
But, I noticed that connection_config doesn't have SSL option. Then, ssl: false is always used as default in presto client library. I would like to add a new parameter to allow users to specify ssl parameter.
For example,

    def connection_config
      {
        catalog: @data_source.dbname,
        server: "#{@data_source.host}:#{@data_source.port}",
        user: @data_source.user,
        password: @data_source.password.presence,
        ssl: {verify: false},
      }.compact
    end

Question

I'm working on making a PR to support ssl option.
But, I noticed that an existing Presto user needs to migrate tables in order to add ssl option if I follow bigquery adaptor implementation. (Ex. I saw data_sources.schema and bigquery_configs.schema)

I want to minimize the code change without any table migration.

Thus, I would like to know whether the following change is acceptable for your team?

• Host parameter allows URL like https://xxxxx/?ssl_verify=false instead of only xxxxx.
• Then, presto_adaptor.rb parses it and build a new connection_config.

If this is acceptable, I'll try to make a PR for this.
If this is not acceptable, I'll follow a way as same as bigquery adaptor.

I'm not familiar with Rails. So, if you have any advise, that would be helpful.

Thank you,

I experienced that some table names are not auto-linked in the table description. e.g. some_log and some_log_latest is auto-linked, but some_log_XXX_latest is not auto-linked. This may be a bug?

Currently, searching is case sensitive. So user may difficult to find schemas.

Currently tables / columns are ordered in random. It should be sorted in alphabetical order (for tables), or in the same order as DB table (for columns).

For example, searching with the word "push" won't list "push_notification_token" table on the result, although other tables like "push_notification_settings" is listed.

I'm not sure about the patterns of table name which will be listed with prefix search or not.

Note: Searching with full table name (not prefix) is working expectedly.

My version is 0.4.0-rc.

Now

• view ... account only
• edit ... account only
• modify datasource ... admin account only

I Want

• view ... everyone
• edit ... account only
• modify datasource ... admin account only

See Also: crowi/crowi#155

Documentation feature for domain-specific keywords is useful.
It is really nice if keywords in table document becomes hyperlink automatically.

Dmemo doesn't work correctly in cluster mode.

• favorite tables
• etc

Maybe 7/3 or 7/4