00:00 - 00:03 | We've received Joshua's uplift plan. |
00:04 - 00:05 | We're going to split the repper logs as they arrive, here |
00:05 - 00:07 | here, and here. |
00:08 - 00:12 | He also thinks we have too many different databases. |
00:12 - 00:15 | He wants to consolidate on just two or three. |
00:17 - 00:19 | That seems like a reasonable plan. |
00:19 - 00:21 | Which ones does he want? |
00:24 - 00:26 | SampleDB is one... |
00:27 - 00:28 | Maybe MySQL... |
00:31 - 00:33 | And HBase. |
00:34 - 00:36 | To replace Sybase IQ. |
00:53 - 00:58 | If you think this is a good idea, get out. You're fired. |
01:13 - 01:15 | He wants to replace Sybase IQ with HBase! |
01:15 - 01:17 | HBase, which crashes at least twice a week! |
01:18 - 01:23 | Replacing IQ, which has been reliable for years, even though it's completely unmaintained! |
01:25 - 01:28 | Did we somehow not explain it well enough? |
01:29 - 01:31 | HBase is completely unusable! |
01:31 - 01:34 | If a mouse farts, the whole thing goes down! |
01:34 - 01:37 | And he thinks this is a suitable replacement for Sybase? |
01:37 - 01:37 | It can't even do OLAP on paper! |
01:40 - 01:42 | He thinks that with an OLAP plug-in -- |
01:42 - 01:46 | Does this plug-in make HBase actually function? |
01:46 - 01:48 | Well, it's still alpha, but -- |
01:48 - 01:52 | Alpha! You want to add an alpha product into the mix! |
01:53 - 01:54 | You want to take the shitpile that is HBase... |
01:56 - 01:57 | ...and then add in an unstable, alpha-level plugin... |
01:57 - 02:00 | ...because HBase by itself isn't unstable enough! |
02:00 - 02:03 | This is a brilliant fucking plan from start to finish! |
02:04 - 02:08 | We take the one part of our architecture that doesn't require constant hand-holding. |
02:08 - 02:13 | The one part that takes whatever we throw at it. |
02:14 - 02:16 | And then we wad it all up and throw it in the fucking trash! |
02:17 - 02:21 | I can't even comprehend the level of stupidity that would think this is a good fucking idea! |
02:27 - 02:29 | Do you even understand how terrible HBase is? |
02:30 - 02:34 | At least twice a week I have to run an hbck because the whole thing is down again. |
02:34 - 02:36 | This is to be our core database! |
02:41 - 02:42 | We're supposed to build all of GTI on top of this! |
02:43 - 02:47 | We're going to use HBase as the fucking bedrock upon which the entire system is built! |
02:48 - 02:53 | You're supposed to use something that's stable as your core technology! |
02:54 - 02:56 | Something that actually works! |
02:56 - 02:59 | But instead we're using the most immature... |
03:00 - 03:02 | ...piece of shit ass-ware in the entire world. |
03:04 - 03:07 | It's okay. At least we're not using MongoDB. |
03:14 - 03:16 | It just doesn't make any sense to me. None at all. |
03:19 - 03:23 | I don't understand why we'd use something that we already know is utter shit. |
03:25 - 03:26 | You'd think we'd at least use something different. |
03:31 - 03:33 | Probably everything else is just as bad. |
03:40 - 03:46 | But at least we'd be able to have a little optimism then. A little hope that things might not go as badly. |
03:46 - 03:49 | The best we can hope for now is that the entire project falls apart. |
03:53 - 03:56 | Then at least we'll fail right away, instead of gradually over the course of a year. |