Troubles 1. Save all of your query’s results AUTOMATICALLY
Troubles 2. hive> alter table tbl drop partition (....); hive> drop table tbl;
Troubles 2. Separate Query executor FROM Administrator’s Client
Troubles 3. hive> set mapred.job.priority=‘VERY_HIGH’ ; hive> SELECT ....
Troubles 3. hive> create temporary function foobar as ‘....’; hive> SELECT foobar(param) ...; FAILED: SemanticException [Error 10011]: Line 1:96 Invalid function ‘foobar’ hive>
Troubles 3. Enforce well-configured environment for almost all of daily queries
Troubles 4. Hey, please count yesterday’s logs. Sure. Wait few minutes... (10 minutes later...) Done. Could you send me by .xlsx? Ah....
Troubles 4. Could you send me by .xlsx? Ah.... OK. Of course. (copy files from HDFS) (open TSV files) (paste all lines on Excel) (open mailer and write e-mail) (send a mail)
Troubles 4. (paste all lines on Excel) (open mailer and write e-mail) (send a mail) (find to miss to attach xlsx) (send a mail again)
Troubles 4. Could you send me by .xlsx? Ah.... OK. Of course. (15 minutes later...) Done!
Troubles 4. (15 minutes later...) Done! Oh, I also need result about the day before yesterday!
Troubles 4. Make all queries shared with permalink and user-friendly formats!
Troubles 4. (fixed) Hey, please count yesterday’s logs. Sure. Wait few minutes... OK. I started query on this URL. So download result after completion. Estimated time is 10 minutes. Thanks!
Shib •Hive Web Client (w/ hiveserver1/2) •Query permalinks •Result datastore (TSV/CSV) •“SELECT” only! (DDL prohibited) •Database/table access control •Query list sharing by tag •HTTP Proxy API (POST to execute query!)
Shib: WebUI tool provides crossover of Hive and MPP
Shib •Multi query engine support •Switch between hiveserver and hiveserver2 •Facebook Presto support •Switch between Hive and Presto!
Shib will supports more features. Enjoy stress-less queries!