<     May 2017     >
Su Mo Tu We Th Fr Sa  
    1  2  3  4  5  6  
 7  8  9 10 11 12 13  
14 15 16 17 18 19 20  
21 22 23 24 25 26 27  
28 _2_9 30 31
02:12 mwlang joined
03:06 GitHub85 joined
03:06 <GitHub85> [13sequel] 15xyhc commented on issue #1350: I alter table varchar(32) to varchar(255) , the issue is fixed. but there is another problem happen: ... 02https://git.io/v9M5X
03:06 GitHub85 left
03:13 GitHub145 joined
03:13 <GitHub145> [13sequel] 15xyhc commented on issue #1350: I alter table varchar(32) to varchar(255) , the issue is fixed. but there is another problem happen: ... 02https://git.io/v9M5X
03:13 GitHub145 left
03:15 GitHub21 joined
03:15 <GitHub21> [13sequel] 15xyhc commented on issue #1350: I alter table varchar(32) to varchar(255) , the issue is fixed. but there is another problem happen: ... 02https://git.io/v9M5X
03:15 GitHub21 left
05:10 glennpratt joined
05:49 GitHub5 joined
05:49 <GitHub5> [13sequel] 15jeremyevans commented on issue #1350: From that error message, it looks like your SQLite database has a foreign key constraint to a table that does not exist. Funnily enough, SQLite allows that as well, though if you enable foreign key support, trying to insert into the table will raise an error. Either remove the foreign key constraint or change it to reference the correct table. 02https://git.io/v9Mp9
05:49 GitHub5 left
05:59 GitHub10 joined
05:59 <GitHub10> [13sequel] 15xyhc commented on issue #1350: OK , thanks a lot ! 02https://git.io/v9MhY
05:59 GitHub10 left
06:13 glennpratt joined
06:14 glennpra_ joined
06:22 glennpratt joined
06:22 glennpratt joined
06:34 glennpratt joined
06:35 glennpratt joined
06:36 glennpratt joined
06:36 glennpratt joined
07:10 glennpratt joined
08:05 glennpratt joined
09:31 GitHub11 joined
09:31 <GitHub11> [13sequel] 15gencer opened issue #1354: Sequel performs bad on rails when used as primary db 02https://git.io/v9DZn
09:31 GitHub11 left
10:42 mmun joined
11:31 glennpratt joined
11:32 GitHub130 joined
11:32 <GitHub130> [13sequel] 15gencer commented on issue #1354: I see that, when `max_connections `parameter provided with **SequelRails**, My app goes down by requests per second. If I remove max_connection it just works as primary.... 02https://git.io/v9DED
11:32 GitHub130 left
12:39 GitHub40 joined
12:39 <GitHub40> [13sequel] 15gencer commented on issue #1354: Yep, this was caused high number of connections per thread. 02https://git.io/v9Do4
12:39 GitHub40 left
12:39 GitHub71 joined
12:39 <GitHub71> [13sequel] 15gencer closed issue #1354: Sequel performs bad on rails when used as primary db without preload_app! parameter and w/ threads 02https://git.io/v9DZn
12:39 GitHub71 left
13:11 glennpra_ joined
13:13 Takumo joined
13:13 Takumo joined
14:06 GitHub48 joined
14:06 <GitHub48> [13sequel] 15jeremyevans commented on issue #1354: This was probably caused because you are creating separate `Sequel::Database` objects in each call to `MasterModel.Model`. Change your initializer to do:... 02https://git.io/v9D5u
14:06 GitHub48 left
14:37 GitHub152 joined
14:37 <GitHub152> [13sequel] 15brauliobo opened issue #1355: Picking up an specific read only slave 02https://git.io/v9Dxq
14:37 GitHub152 left
14:50 GitHub154 joined
14:50 <GitHub154> [13sequel] 15gencer commented on issue #1354: Oh, this is awesome. Thanks. Let me test this now.... 02https://git.io/v9Dhc
14:50 GitHub154 left
14:52 GitHub193 joined
14:52 <GitHub193> [13sequel] 15gencer commented on issue #1354: Oh, this is awesome. Thanks. Let me test this now.... 02https://git.io/v9Dhc
14:52 GitHub193 left
15:21 GitHub149 joined
15:21 <GitHub149> [13sequel] 15gencer commented on issue #1354: @jeremyevans, I tested your solution. It does not work with high number of max_connection.... 02https://git.io/v9yT8
15:21 GitHub149 left
15:25 GitHub196 joined
15:25 <GitHub196> [13sequel] 15gencer commented on issue #1354: @jeremyevans, I tested your solution. It does not work with high number of max_connection.... 02https://git.io/v9yT8
15:25 GitHub196 left
15:31 GitHub33 joined
15:31 <GitHub33> [13sequel] 15gencer commented on issue #1354: @jeremyevans one more question. Could be this slowness be happen because of pgpool2 extension? Currently, I connect to pgpool and then to postgresql.... 02https://git.io/v9yIS
15:31 GitHub33 left
16:35 GitHub134 joined
16:35 <GitHub134> [13sequel] 15jeremyevans commented on issue #1354: 24 workers * 100 connections is 2400 connections, which is probably taxing your database server way too much. I would stay at 4 connections/worker, and only increase if you have benchmarked and found a higher number to be faster. Instead of testing at 100, try testing at 6 or 8 and seeing what the differences are compared to 4.... 02https://git.io/v9ynN
16:35 GitHub134 left
16:48 GitHub156 joined
16:48 <GitHub156> [13sequel] 15jeremyevans commented on issue #1355: If you use `with_server`, you are telling Sequel that you want database activity inside the block to go to that server unless the dataset specifically overrides the server to use. So if you don't want the `INSERT` to go to the `api_customers` shard, then you shouldn't use `with_server(:api_customers)`... 02https://git.io/v9yCh
16:48 GitHub156 left
16:48 GitHub29 joined
16:48 <GitHub29> [13sequel] 15jeremyevans closed issue #1355: Picking up an specific read only slave 02https://git.io/v9Dxq
16:48 GitHub29 left
16:52 GitHub127 joined
16:52 <GitHub127> [13sequel] 15StevenCregan commented on issue #1351: Using :database_type == :oracle didn't work for me either 02https://git.io/v9ylk
16:52 GitHub127 left
16:55 GitHub131 joined
16:55 <GitHub131> [13sequel] 15brauliobo commented on issue #1355: When using without the `with_server` I don't need to worry about if it is an `INSERT` or an `SELECT`. So what I need is a way to specify a read replica and keep this smart behaviour. 02https://git.io/v9ylp
16:55 GitHub131 left
17:27 glennpratt joined
17:33 glennpratt joined
17:35 GitHub145 joined
17:35 <GitHub145> [13sequel] 15jeremyevans commented on issue #1355: Unfortunately, that is not currently possible with the server_block extension. You would have to modify the server_block extension or add a separate extension to support it. 02https://git.io/v9ygg
17:35 GitHub145 left
17:36 glennpratt joined
17:38 GitHub103 joined
17:38 <GitHub103> [13sequel] 15brauliobo commented on issue #1355: Ok, that is also not possible with the `server` call, so I presume that problem isn't on the `server_block` extension 02https://git.io/v9ygH
17:38 GitHub103 left
17:38 GitHub152 joined
17:38 <GitHub152> [13sequel] 15brauliobo commented on issue #1355: Ok, that is also not possible with the `server` call, so I presume that problem isn't on the `server_block` extension 02https://git.io/v9ygH
17:38 GitHub152 left
17:38 glennpratt joined
17:40 GitHub46 joined
17:40 <GitHub46> [13sequel] 15jeremyevans commented on issue #1355: You could always call `Dataset#server` to specify the server to use for each dataset (separately for the INSERT and SELECT). However, in general Sequel uses the same server for the SELECT after the INSERT in a Model.create call. Otherwse, there is often a race condition where the read replica won't have have received the row before Sequel tries to select from it. 02https://git.io/v9y2Y
17:40 GitHub46 left
17:42 GitHub51 joined
17:42 <GitHub51> [13sequel] 15brauliobo commented on issue #1355: @jeremyevans that is the big hassle: in large applications it is completely inviable to spread calls of `Dataset#server`, also considering that the server parameter would need to be passed to several contexts 02https://git.io/v9y21
17:42 GitHub51 left
17:46 GitHub35 joined
17:46 <GitHub35> [13sequel] 15jeremyevans commented on issue #1355: As the contribution guidelines mention, GitHub Issues is not a support forum. This isn't a bug in Sequel. If you'd like it to be considered as an additional feature (specify regular and read_only shards in with_server), please post on the sequel-talk Google Group and see if there is any community support for that. 02https://git.io/v9ya4
17:46 GitHub35 left
17:46 GitHub26 joined
17:46 <GitHub26> [13sequel] 15brauliobo commented on issue #1355: Wouldn't it be simple to add a readonly flag to the server hash? So it would be easy for sequel to know if it is a master or not. Also would be interesting to have the read only replicas inside their respective *master*. As it is now, there is no way to know about which replica is for which master. 02https://git.io/v9yaz
17:46 GitHub26 left
17:47 GitHub160 joined
17:47 <GitHub160> [13sequel] 15brauliobo commented on issue #1355: Sorry @jeremyevans, I seriously think this is a limitation, so this issue is a feature request IHMO 02https://git.io/v9yar
17:47 GitHub160 left
17:50 GitHub103 joined
17:50 <GitHub103> [13sequel] 15jeremyevans commented on issue #1355: GitHub Issues should not be used for feature requests, only for bugs, as specified in the contribution guidelines. You can submit a pull request for features. 02https://git.io/v9yVq
17:50 GitHub103 left
17:50 glennpratt joined
17:52 GitHub104 joined
17:52 <GitHub104> [13sequel] 15brauliobo commented on issue #1355: thanks @jeremyevans, I consider some time to improve this! 02https://git.io/v9yVD
17:52 GitHub104 left
17:53 GitHub4 joined
17:53 <GitHub4> [13sequel] 15brauliobo commented on issue #1355: thanks @jeremyevans, I consider some time to improve this! 02https://git.io/v9yVD
17:53 GitHub4 left
17:54 glennpra_ joined
18:02 GitHub25 joined
18:02 <GitHub25> [13sequel] 15gencer commented on issue #1354: @jeremyevans thanks for the explanation. Yes, you're right. I also made a very simple mistake. I did 2 different app to connect same database and do pooling on it. Due to this it just goes down. I seperated apps, lowered threads and used even high number of max_connections (higher than 4) for each one, dropped pgpool and directly connected to postgresql's 5432 port. It just works. I do stress tests. Even
18:02 GitHub25 left
20:26 tercenya joined
22:20 tercenya joined