Commit Graph

254 Commits

Author SHA1 Message Date
4f8f4b96cd Fix docker_enabled inconsistency 2025-11-14 23:41:16 +01:00
7544fbb1ce Refactor JSON marshaling in Options to improve thread safety 2025-11-14 21:50:58 +01:00
4f4feacaa8 Remove manual config fields assignment 2025-11-14 20:38:36 +01:00
2c28971e15 Add JSON tags to configuration structs 2025-11-14 20:37:33 +01:00
e77ed0cdef Deep copy config for sanitization 2025-11-14 20:25:40 +01:00
623e258a2a Add API endpoint to retrieve sanitized server configuration 2025-11-14 18:57:03 +01:00
511889e56d Implement per instance command override on backend 2025-11-14 18:38:31 +01:00
11bfe75a3c Add support for extra args for command parser 2025-11-13 20:41:08 +01:00
ae5358ff65 Change FlashAttn field type to string in LlamaServerOptions 2025-11-12 23:49:34 +01:00
bff8e7d914 Refactor command line argument keys to use snake_case format in LlamaServerOptions 2025-11-12 23:46:15 +01:00
5ad076919e Add extra args test to backend 2025-11-12 23:40:07 +01:00
8d92f9b371 Add ExtraArgs support for Llama, Mlx, and Vllm server options 2025-11-12 23:11:22 +01:00
560850f86d Add shutdown state checks in HTTP handlers 2025-10-30 18:00:59 +01:00
77c0e22fd0 Use instance's ServeHTTP in handlers 2025-10-29 00:01:29 +01:00
d65c5ab717 Wait for inflight requests before stopping 2025-10-29 00:00:56 +01:00
2b94244c8a Replace GetProxy with ServeHttp in instance 2025-10-29 00:00:02 +01:00
2e5644db53 Implement inflight request tracking 2025-10-28 23:59:02 +01:00
7ee22fee51 Implement shutting down status 2025-10-28 23:53:11 +01:00
f9eb424690 Fix concurrent map write issue in MarshalJSON by initializing BackendOptions 2025-10-27 20:36:42 +01:00
dfcc16083c Update test configuration to use 'sh -c "sleep 999999"' command 2025-10-27 19:25:13 +01:00
6ec2919049 Fix instance start simulation in TestUpdateInstance 2025-10-27 19:14:54 +01:00
d6a6f377fc Fix logger race condition 2025-10-27 19:06:06 +01:00
cd9a71d9fc Update test configuration to use 'yes' command instead of 'sleep' 2025-10-27 18:54:20 +01:00
2c4cc5a69a Fix manager tests 2025-10-27 18:47:17 +01:00
b1fc1d2dc8 Add InstancesDir to test configuration for instance management 2025-10-27 18:38:23 +01:00
08c47a16a0 Fix operations tests 2025-10-27 18:35:16 +01:00
219db7abce Move port range validation to config 2025-10-27 18:23:49 +01:00
14131a6274 Remove redundant code 2025-10-27 18:18:25 +01:00
e65f4f1641 Remove unsupported error wrapping from log.Printf 2025-10-27 18:01:58 +01:00
5ef0654cdd Use %w for error wrapping in log messages across multiple files 2025-10-27 17:54:39 +01:00
1814772fa2 Fix instance deletion check to account for restarting status 2025-10-27 17:42:27 +01:00
f94d05dad2 Add Restarting state 2025-10-26 18:55:05 +01:00
249ff2a7aa Capitalize godoc tags 2025-10-26 16:49:27 +01:00
59c954811d Update API routes in godoc 2025-10-26 16:35:42 +01:00
58c8899fd9 Update import path for API documentation 2025-10-26 14:08:48 +01:00
f98b09ea78 Move apidocs to docs folder 2025-10-26 14:04:53 +01:00
969fee837f Fix instance name retrieval 2025-10-26 11:34:45 +01:00
4e587953d8 Refactor llama server command handlers to use a common execution function 2025-10-26 11:00:10 +01:00
356c5be2c6 Improve comments 2025-10-26 10:34:36 +01:00
836e918fc5 Rename ProxyToInstance to InstanceProxy for clarity in routing 2025-10-26 10:22:37 +01:00
a7593e9a58 Split LlamaCppProxy handler 2025-10-26 10:21:40 +01:00
9259763054 Add getInstance method to handlers 2025-10-26 09:54:24 +01:00
94dce4c9bb Implement helper response handling functions 2025-10-26 00:12:33 +02:00
a3f9213f04 Implement ensureInstanceRunning helper 2025-10-25 23:44:21 +02:00
de5a38e7fd Refactor command parsing 2025-10-25 20:23:08 +02:00
ea6c76cc96 Update multi valued flags in backends 2025-10-25 19:02:46 +02:00
bd6436840e Implement common ParseCommand interface 2025-10-25 18:41:46 +02:00
c038aac91b Remove redundant UpdateLast RequestTime calls 2025-10-25 16:09:57 +02:00
7d9b983f93 Don't strip remote llama-cpp proxy prefix 2025-10-25 16:02:09 +02:00
889df3cb79 Add API key header for remote instances in proxy build 2025-10-25 14:14:39 +02:00