[install-help]: Is ad blocker stopping context chat?

[knot-split]

Describe the issue
The backend is on one computer and the next cloud installation is on another. They are successfully communicating by way of the harp proxy But there are a few glitches I'm trying to work out. Sometimes the source loading process stalls and I have to restart the container for the backend and the front-end enabled/disabled in order to get moving again. After trial and error, my theory tells me that my pi-hole ad-blocker somehow prevents this loading process. So I've been turning it off and I find that the stall doesn't seem to persist as long.

We haven't yet passed the initial indexing, but it seems we made tremendous progress.

My question is how to configure the pi hole so that I can keep the ad blocker and not tamper the context chat backend adversely.

Context Chat Backend logs (if applicable, from the docker container)

Details

``` 2026-04-18 09:28:16.064 | INFO: - "PUT /loadSources HTTP/1.1" 200 OK 2026-04-18 09:28:20.480 | INFO: - "PUT /loadSources HTTP/1.1" 200 OK 2026-04-18 09:28:33.552 | INFO: - "PUT /loadSources HTTP/1.1" 200 OK ```

Setup Details (please complete the following information):

• Nextcloud Version: AIO current
• AppAPI Version: current
• Context Chat PHP Version 5.3.1
• Context Chat Backend Version 5.3. dev
• Nextcloud deployment method: AIO
• Context Chat Backend deployment method: manual

[kyteinsky]

hello,
not quite sure how pihole could interfere with these requests.
Would you mind providing some more insight into the systems, it may be a RAM issue, network configuration issue since it's a remote ex-app server, and might be DNS if the remote machine is addressed by a hostname which is where pihole could be doing it's funny business.
Also, please post the harp daemon config details.

[knot-split]

Here's some system info:

Linux e 6.6.87.2-microsoft-standard-WSL2 #1 SMP PREEMPT_DYNAMIC Thu Jun 5 18:30:46 UTC 2025 x86_64 x86_64 x86_64 GNU/Linux

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 595.58.02              Driver Version: 595.97         CUDA Version: 13.2     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 5070 Ti     On  |   00000000:01:00.0  On |                  N/A |
|  0%   33C    P8             11W /  300W |    8412MiB /  16303MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

Here's a log from Context Chat App

`{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:28+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"Background jobs: OCA\\ContextChat\\BackgroundJobs\\IndexerJob 69801 started","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:28+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"[IndexerJob] Running indexing","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":{"storageId":"3","rootId":"191"}}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:29+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"[IndexerJob] Indexed 1 files","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":{"storageId":"3","rootId":"191","loadedSources":"[\"files__default: 224950\"]","sourcesToRetry":"[]","timeTaken":"1","totalSize":"69","sources":"[\"224950:/christopher/files/Notes/A piano and a penguin adjacent, both standing on a checkerboard floor.md\"]"}}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:29+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"Indexed 1 files","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:29+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"[IndexerJob] Batch processed","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":{"storageId":"3","rootId":"191","nFilesProcessed":"1","nFilesEligible":"1","nFilesLoaded":"1","nFilesToRetry":"0","nFilesInvalidOrEmpty":"0","nRetryQFiles":"0","filesProcessed":"[\"224950:/christopher/files/Notes/A piano and a penguin adjacent, both standing on a checkerboard floor.md\"]","filesEligible":"[\"224950:/christopher/files/Notes/A piano and a penguin adjacent, both standing on a checkerboard floor.md\"]","filesLoaded":"[\"224950:/christopher/files/Notes/A piano and a penguin adjacent, both standing on a checkerboard floor.md\"]","filesToRetry":"[]","filesInvalidOrEmpty":"[]","retryQFiles":"[]"}}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:29+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"[IndexerJob]  Removing OCA\\ContextChat\\BackgroundJobs\\IndexerJob with argument array (\n  'storageId' => 3,\n  'rootId' => 191,\n)from oc_jobs","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}
{"reqId":"EizW9CgqnusTxoTErbfM","level":3,"time":"2026-04-18T23:01:29+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"Background jobs: OCA\\ContextChat\\BackgroundJobs\\IndexerJob 69801 ended","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}
{"reqId":"InbBLsZvvYYtZpALSs4L","level":3,"time":"2026-04-18T23:06:28+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"Background jobs: OCA\\ContextChat\\BackgroundJobs\\FileSystemListenerJob 69676 started","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}
{"reqId":"InbBLsZvvYYtZpALSs4L","level":3,"time":"2026-04-18T23:06:28+00:00","remoteAddr":"","user":"--","app":"no app in context","method":"","url":"--","scriptName":"/var/www/html/cron.php","message":"Background jobs: OCA\\ContextChat\\BackgroundJobs\\FileSystemListenerJob 69676 ended","userAgent":"--","version":"32.0.8.2","occ_command":["/var/www/html/cron.php"],"data":[]}`

specs from nvidia app:

NVIDIA system information report created on: 04/21/2026 15:32:23
NVIDIA App version: 11.0.7.237
Operating system: Microsoft Windows 11 Pro, Version 10.0.26200
DirectX runtime version: DirectX 12
Driver: Game Ready Driver - 595.97 - Tue Mar 24, 2026
CPU: 12th Gen Intel(R) Core(TM) i7-12700
RAM: 64.0 GB
Storage (4): SSD - 931.5 GB,HDD - 931.5 GB,+2 more

Here's a typcial backend startup....I'm also thinking now that the indexing process gets about halfway through 5000 files and then it just gets stuck. Anyhow, here's the startup info:

 /usr/lib/python3/dist-packages/supervisor/options.py:13: UserWarning: pkg_resources is deprecated as an API. See https://setuptools.pypa.io/en/latest/pkg_resources.html. The pkg_resources package is slated for removal as early as 2025-11-30. Refrain from using this package or pin to Setuptools<81.
2026-04-18 08:22:22.232 |   import pkg_resources
2026-04-18 08:22:23.338 | HP_SHARED_KEY is set, creating /frpc.toml configuration file...
2026-04-18 08:22:23.338 | Found /certs/frp directory. Creating configuration with TLS certificates.
2026-04-18 08:22:23.339 | Starting frpc...
2026-04-18 08:22:23.357 | Starting PostgreSQL...
2026-04-18 08:22:23.358 | 2026-04-18 13:22:23.357 [I] [sub/root.go:142] start frpc service for config file [/frpc.toml]
2026-04-18 08:22:23.358 | 2026-04-18 13:22:23.357 [I] [client/service.go:295] try to connect to server...
2026-04-18 08:22:23.377 | pg_ctl: another server might be running; trying to start server anyway
2026-04-18 08:22:23.548 | waiting for server to start....2026-04-18 13:22:23.548 [I] [client/service.go:287] [0192f6db9852df94] login to server success, get run id [0192f6db9852df94]
2026-04-18 08:22:23.549 | 2026-04-18 13:22:23.548 [I] [proxy/proxy_manager.go:173] [0192f6db9852df94] proxy added: [context_chat_backend]
2026-04-18 08:22:23.551 | 2026-04-18 13:22:23.551 [I] [client/control.go:168] [0192f6db9852df94] [context_chat_backend] start proxy success
2026-04-18 08:22:23.700 |  done
2026-04-18 08:22:23.700 | server started
2026-04-18 08:22:23.701 | Waiting for PostgreSQL to start...
2026-04-18 08:22:23.741 | PostgreSQL is up and running.
2026-04-18 08:22:23.741 | CCB_DB_URL is already set. Skipping database setup.
2026-04-18 08:22:27.494 | Waiting for 10 seconds before starting embedding server to allow main app to start
2026-04-18 08:22:27.914 | Running repairs...
2026-04-18 08:22:27.914 | No repairs are required.
2026-04-18 08:22:27.917 | Config file already exists in the persistent storage ("persistent_storage/config.yaml").
2026-04-18 08:22:27.938 | App config:
2026-04-18 08:22:27.938 | {
2026-04-18 08:22:27.938 |   "debug": true,
2026-04-18 08:22:27.938 |   "uvicorn_log_level": "info",
2026-04-18 08:22:27.938 |   "disable_aaa": false,
2026-04-18 08:22:27.938 |   "verify_ssl": true,
2026-04-18 08:22:27.938 |   "use_colors": true,
2026-04-18 08:22:27.938 |   "uvicorn_workers": 1,
2026-04-18 08:22:27.938 |   "embedding_chunk_size": 2000,
2026-04-18 08:22:27.938 |   "doc_parser_worker_limit": 10,
2026-04-18 08:22:27.938 |   "vectordb": [
2026-04-18 08:22:27.938 |     "pgvector",
2026-04-18 08:22:27.938 |     {}
2026-04-18 08:22:27.938 |   ],
2026-04-18 08:22:27.938 |   "embedding": {
2026-04-18 08:22:27.938 |     "base_url": "http://localhost:5000/v1",
2026-04-18 08:22:27.938 |     "workers": 1,
2026-04-18 08:22:27.938 |     "request_timeout": 1800,
2026-04-18 08:22:27.938 |     "model_name": "Qwen3-8B-Jailbroken.Q4_K_M.gguf",
2026-04-18 08:22:27.938 |     "auth": null,
2026-04-18 08:22:27.938 |     "remote_service": false,
2026-04-18 08:22:27.938 |     "batch_size": 100,
2026-04-18 08:22:27.938 |     "llama": {
2026-04-18 08:22:27.938 |       "model": "granite-embedding-107m-multilingual-f16.gguf",
2026-04-18 08:22:27.938 |       "n_batch": 16,
2026-04-18 08:22:27.938 |       "n_ctx": 8192,
2026-04-18 08:22:27.938 |       "n_gpu_layers": 33
2026-04-18 08:22:27.938 |     }
2026-04-18 08:22:27.938 |   },
2026-04-18 08:22:27.938 |   "llm": [
2026-04-18 08:22:27.938 |     "nc_texttotext",
2026-04-18 08:22:27.938 |     {}
2026-04-18 08:22:27.938 |   ]
2026-04-18 08:22:27.938 | }
2026-04-18 08:22:27.967 | INFO:     Started server process [41]
2026-04-18 08:22:27.967 | INFO:     Waiting for application startup.
2026-04-18 08:22:28.451 | INFO:     Application startup complete.
2026-04-18 08:22:28.452 | INFO:     Uvicorn running on unix socket /tmp/exapp.sock (Press CTRL+C to quit)
2026-04-18 08:22:39.583 | Config file already exists in the persistent storage ("persistent_storage/config.yaml").
2026-04-18 08:22:39.596 | Embedder config:
2026-04-18 08:22:39.596 | {
2026-04-18 08:22:39.596 |   "base_url": "http://localhost:5000/v1",
2026-04-18 08:22:39.596 |   "workers": 1,
2026-04-18 08:22:39.596 |   "request_timeout": 1800,
2026-04-18 08:22:39.596 |   "model_name": "Qwen3-8B-Jailbroken.Q4_K_M.gguf",
2026-04-18 08:22:39.596 |   "auth": null,
2026-04-18 08:22:39.596 |   "remote_service": false,
2026-04-18 08:22:39.596 |   "batch_size": 100,
2026-04-18 08:22:39.596 |   "llama": {
2026-04-18 08:22:39.596 |     "model": "granite-embedding-107m-multilingual-f16.gguf",
2026-04-18 08:22:39.596 |     "n_batch": 16,
2026-04-18 08:22:39.596 |     "n_ctx": 8192,
2026-04-18 08:22:39.596 |     "n_gpu_layers": 33
2026-04-18 08:22:39.596 |   }
2026-04-18 08:22:39.596 | }
2026-04-18 08:22:39.608 | INFO:      - "GET /enabled HTTP/1.1" 200 OK
2026-04-18 08:22:49.882 | ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    yes
2026-04-18 08:22:49.882 | ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
2026-04-18 08:22:49.882 | ggml_cuda_init: found 1 CUDA devices:
2026-04-18 08:22:49.882 |   Device 0: NVIDIA GeForce RTX 5070 Ti, compute capability 12.0, VMM: yes
2026-04-18 08:22:50.075 | llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 5070 Ti) - 15011 MiB free
2026-04-18 08:22:50.110 | llama_model_loader: loaded meta data with 37 key-value pairs and 101 tensors from persistent_storage/model_files/granite-embedding-107m-multilingual-f16.gguf (version GGUF V3 (latest))
2026-04-18 08:22:50.110 | llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
2026-04-18 08:22:50.110 | llama_model_loader: - kv   0:                       general.architecture str              = bert
2026-04-18 08:22:50.110 | llama_model_loader: - kv   1:                               general.type str              = model
2026-04-18 08:22:50.110 | llama_model_loader: - kv   2:                               general.name str              = Granite Embedding 107m Multilingual
2026-04-18 08:22:50.110 | llama_model_loader: - kv   3:                           general.finetune str              = multilingual
2026-04-18 08:22:50.110 | llama_model_loader: - kv   4:                           general.basename str              = granite-embedding
2026-04-18 08:22:50.110 | llama_model_loader: - kv   5:                         general.size_label str              = 107M
2026-04-18 08:22:50.110 | llama_model_loader: - kv   6:                            general.license str              = apache-2.0
2026-04-18 08:22:50.110 | llama_model_loader: - kv   7:                               general.tags arr[str,5]       = ["language", "granite", "embeddings",...
2026-04-18 08:22:50.110 | llama_model_loader: - kv   8:                          general.languages arr[str,12]      = ["en", "ar", "cs", "de", "es", "fr", ...
2026-04-18 08:22:50.110 | llama_model_loader: - kv   9:                           bert.block_count u32              = 6
2026-04-18 08:22:50.110 | llama_model_loader: - kv  10:                        bert.context_length u32              = 512
2026-04-18 08:22:50.110 | llama_model_loader: - kv  11:                      bert.embedding_length u32              = 384
2026-04-18 08:22:50.110 | llama_model_loader: - kv  12:                   bert.feed_forward_length u32              = 1536
2026-04-18 08:22:50.110 | llama_model_loader: - kv  13:                  bert.attention.head_count u32              = 12
2026-04-18 08:22:50.110 | llama_model_loader: - kv  14:          bert.attention.layer_norm_epsilon f32              = 0.000000
2026-04-18 08:22:50.110 | llama_model_loader: - kv  15:                          general.file_type u32              = 1
2026-04-18 08:22:50.110 | llama_model_loader: - kv  16:                      bert.attention.causal bool             = false
2026-04-18 08:22:50.110 | llama_model_loader: - kv  17:                          bert.pooling_type u32              = 2
2026-04-18 08:22:50.110 | llama_model_loader: - kv  18:                       tokenizer.ggml.model str              = t5
2026-04-18 08:22:50.110 | llama_model_loader: - kv  19:                         tokenizer.ggml.pre str              = default
2026-04-18 08:22:50.133 | llama_model_loader: - kv  20:                      tokenizer.ggml.tokens arr[str,250002]  = ["<s>", "<pad>", "</s>", "<unk>", ","...
2026-04-18 08:22:50.184 | llama_model_loader: - kv  21:                      tokenizer.ggml.scores arr[f32,250002]  = [0.000000, 0.000000, 0.000000, 0.0000...
2026-04-18 08:22:50.188 | llama_model_loader: - kv  22:                  tokenizer.ggml.token_type arr[i32,250002]  = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ...
2026-04-18 08:22:50.188 | llama_model_loader: - kv  23:            tokenizer.ggml.add_space_prefix bool             = true
2026-04-18 08:22:50.188 | llama_model_loader: - kv  24:            tokenizer.ggml.token_type_count u32              = 2
2026-04-18 08:22:50.188 | llama_model_loader: - kv  25:    tokenizer.ggml.remove_extra_whitespaces bool             = true
2026-04-18 08:22:50.193 | llama_model_loader: - kv  26:        tokenizer.ggml.precompiled_charsmap arr[u8,237539]   = [0, 180, 2, 0, 0, 132, 0, 0, 0, 0, 0,...
2026-04-18 08:22:50.193 | llama_model_loader: - kv  27:                tokenizer.ggml.bos_token_id u32              = 0
2026-04-18 08:22:50.193 | llama_model_loader: - kv  28:                tokenizer.ggml.eos_token_id u32              = 2
2026-04-18 08:22:50.193 | llama_model_loader: - kv  29:            tokenizer.ggml.unknown_token_id u32              = 3
2026-04-18 08:22:50.193 | llama_model_loader: - kv  30:          tokenizer.ggml.seperator_token_id u32              = 2
2026-04-18 08:22:50.193 | llama_model_loader: - kv  31:            tokenizer.ggml.padding_token_id u32              = 1
2026-04-18 08:22:50.193 | llama_model_loader: - kv  32:                tokenizer.ggml.cls_token_id u32              = 0
2026-04-18 08:22:50.193 | llama_model_loader: - kv  33:               tokenizer.ggml.mask_token_id u32              = 250001
2026-04-18 08:22:50.193 | llama_model_loader: - kv  34:               tokenizer.ggml.add_bos_token bool             = true
2026-04-18 08:22:50.193 | llama_model_loader: - kv  35:               tokenizer.ggml.add_eos_token bool             = true
2026-04-18 08:22:50.193 | llama_model_loader: - kv  36:               general.quantization_version u32              = 2
2026-04-18 08:22:50.193 | llama_model_loader: - type  f32:   64 tensors
2026-04-18 08:22:50.193 | llama_model_loader: - type  f16:   37 tensors
2026-04-18 08:22:50.193 | print_info: file format = GGUF V3 (latest)
2026-04-18 08:22:50.193 | print_info: file type   = F16
2026-04-18 08:22:50.193 | print_info: file size   = 204.23 MiB (16.03 BPW) 
2026-04-18 08:22:50.244 | init_tokenizer: initializing tokenizer for type 4
2026-04-18 08:22:50.406 | load: model vocab missing newline token, using special_pad_id instead
2026-04-18 08:22:50.478 | load: control token:      2 '</s>' is not marked as EOG
2026-04-18 08:22:50.479 | load: control token:      1 '<pad>' is not marked as EOG
2026-04-18 08:22:50.482 | load: control token:      0 '<s>' is not marked as EOG
2026-04-18 08:22:50.493 | load: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
2026-04-18 08:22:50.493 | load: special tokens cache size = 4
2026-04-18 08:22:50.517 | load: token to piece cache size = 2.1668 MB
2026-04-18 08:22:50.517 | print_info: arch             = bert
2026-04-18 08:22:50.517 | print_info: vocab_only       = 0
2026-04-18 08:22:50.517 | print_info: n_ctx_train      = 512
2026-04-18 08:22:50.517 | print_info: n_embd           = 384
2026-04-18 08:22:50.517 | print_info: n_layer          = 6
2026-04-18 08:22:50.517 | print_info: n_head           = 12
2026-04-18 08:22:50.517 | print_info: n_head_kv        = 12
2026-04-18 08:22:50.517 | print_info: n_rot            = 32
2026-04-18 08:22:50.517 | print_info: n_swa            = 0
2026-04-18 08:22:50.517 | print_info: is_swa_any       = 0
2026-04-18 08:22:50.517 | print_info: n_embd_head_k    = 32
2026-04-18 08:22:50.517 | print_info: n_embd_head_v    = 32
2026-04-18 08:22:50.517 | print_info: n_gqa            = 1
2026-04-18 08:22:50.517 | print_info: n_embd_k_gqa     = 384
2026-04-18 08:22:50.517 | print_info: n_embd_v_gqa     = 384
2026-04-18 08:22:50.517 | print_info: f_norm_eps       = 1.0e-12
2026-04-18 08:22:50.517 | print_info: f_norm_rms_eps   = 0.0e+00
2026-04-18 08:22:50.517 | print_info: f_clamp_kqv      = 0.0e+00
2026-04-18 08:22:50.517 | print_info: f_max_alibi_bias = 0.0e+00
2026-04-18 08:22:50.517 | print_info: f_logit_scale    = 0.0e+00
2026-04-18 08:22:50.517 | print_info: f_attn_scale     = 0.0e+00
2026-04-18 08:22:50.517 | print_info: n_ff             = 1536
2026-04-18 08:22:50.517 | print_info: n_expert         = 0
2026-04-18 08:22:50.517 | print_info: n_expert_used    = 0
2026-04-18 08:22:50.517 | print_info: causal attn      = 0
2026-04-18 08:22:50.517 | print_info: pooling type     = 2
2026-04-18 08:22:50.517 | print_info: rope type        = 2
2026-04-18 08:22:50.517 | print_info: rope scaling     = linear
2026-04-18 08:22:50.517 | print_info: freq_base_train  = 10000.0
2026-04-18 08:22:50.517 | print_info: freq_scale_train = 1
2026-04-18 08:22:50.517 | print_info: n_ctx_orig_yarn  = 512
2026-04-18 08:22:50.517 | print_info: rope_finetuned   = unknown
2026-04-18 08:22:50.517 | print_info: model type       = 22M
2026-04-18 08:22:50.517 | print_info: model params     = 106.85 M
2026-04-18 08:22:50.517 | print_info: general.name     = Granite Embedding 107m Multilingual
2026-04-18 08:22:50.517 | print_info: vocab type       = UGM
2026-04-18 08:22:50.517 | print_info: n_vocab          = 250002
2026-04-18 08:22:50.517 | print_info: n_merges         = 0
2026-04-18 08:22:50.517 | print_info: BOS token        = 0 '<s>'
2026-04-18 08:22:50.517 | print_info: EOS token        = 2 '</s>'
2026-04-18 08:22:50.517 | print_info: UNK token        = 3 '<unk>'
2026-04-18 08:22:50.517 | print_info: SEP token        = 2 '</s>'
2026-04-18 08:22:50.517 | print_info: PAD token        = 1 '<pad>'
2026-04-18 08:22:50.517 | print_info: MASK token       = 250001 '[PAD250000]'
2026-04-18 08:22:50.517 | print_info: LF token         = 0 '<s>'
2026-04-18 08:22:50.517 | print_info: EOG token        = 2 '</s>'
2026-04-18 08:22:50.517 | print_info: max token length = 48
2026-04-18 08:22:50.518 | load_tensors: loading model tensors, this can take a while... (mmap = true)
2026-04-18 08:22:50.518 | load_tensors: layer   0 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   1 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   2 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   3 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   4 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   5 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.518 | load_tensors: layer   6 assigned to device CUDA0, is_swa = 0
2026-04-18 08:22:50.519 | load_tensors: tensor 'token_embd.weight' (f16) (and 4 others) cannot be used with preferred buffer type CUDA_Host, using CPU instead
2026-04-18 08:22:50.571 | load_tensors: offloading 6 repeating layers to GPU
2026-04-18 08:22:50.571 | load_tensors: offloading output layer to GPU
2026-04-18 08:22:50.571 | load_tensors: offloaded 7/7 layers to GPU
2026-04-18 08:22:50.571 | load_tensors:        CUDA0 model buffer size =    20.36 MiB
2026-04-18 08:22:50.571 | load_tensors:   CPU_Mapped model buffer size =   183.86 MiB
2026-04-18 08:22:50.586 | ...........
2026-04-18 08:22:50.587 | llama_context: constructing llama_context
2026-04-18 08:22:50.587 | llama_context: n_batch is less than GGML_KQ_MASK_PAD - increasing to 64
2026-04-18 08:22:50.587 | llama_context: n_seq_max     = 1
2026-04-18 08:22:50.587 | llama_context: n_ctx         = 8192
2026-04-18 08:22:50.587 | llama_context: n_ctx_per_seq = 8192
2026-04-18 08:22:50.587 | llama_context: n_batch       = 64
2026-04-18 08:22:50.587 | llama_context: n_ubatch      = 16
2026-04-18 08:22:50.587 | llama_context: causal_attn   = 0
2026-04-18 08:22:50.587 | llama_context: flash_attn    = 0
2026-04-18 08:22:50.587 | llama_context: freq_base     = 10000.0
2026-04-18 08:22:50.587 | llama_context: freq_scale    = 1
2026-04-18 08:22:50.587 | llama_context: n_ctx_per_seq (8192) > n_ctx_train (512) -- possible training context overflow
2026-04-18 08:22:50.587 | set_abort_callback: call
2026-04-18 08:22:50.589 | llama_context:  CUDA_Host  output buffer size =     0.96 MiB
2026-04-18 08:22:50.589 | llama_context: enumerating backends
2026-04-18 08:22:50.589 | llama_context: backend_ptrs.size() = 2
2026-04-18 08:22:50.589 | llama_context: max_nodes = 65536
2026-04-18 08:22:50.600 | CUDA : ARCHS = 500,520,530,600,610,620,700,720,750,800,860,870,890,900 | FORCE_MMQ = 1 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | AVX2 = 1 | F16C = 1 | FMA = 1 | BMI2 = 1 | LLAMAFILE = 1 | OPENMP = 1 | REPACK = 1 | 
2026-04-18 08:22:50.600 | Model metadata: {'general.quantization_version': '2', 'tokenizer.ggml.add_eos_token': 'true', 'bert.embedding_length': '384', 'bert.block_count': '6', 'general.architecture': 'bert', 'bert.feed_forward_length': '1536', 'bert.context_length': '512', 'general.license': 'apache-2.0', 'general.type': 'model', 'bert.pooling_type': '2', 'bert.attention.layer_norm_epsilon': '0.000000', 'bert.attention.causal': 'false', 'tokenizer.ggml.pre': 'default', 'general.basename': 'granite-embedding', 'tokenizer.ggml.seperator_token_id': '2', 'tokenizer.ggml.padding_token_id': '1', 'general.finetune': 'multilingual', 'general.file_type': '1', 'tokenizer.ggml.model': 't5', 'tokenizer.ggml.add_space_prefix': 'true', 'bert.attention.head_count': '12', 'tokenizer.ggml.token_type_count': '2', 'tokenizer.ggml.cls_token_id': '0', 'tokenizer.ggml.remove_extra_whitespaces': 'true', 'general.size_label': '107M', 'tokenizer.ggml.add_bos_token': 'true', 'general.name': 'Granite Embedding 107m Multilingual', 'tokenizer.ggml.bos_token_id': '0', 'tokenizer.ggml.eos_token_id': '2', 'tokenizer.ggml.unknown_token_id': '3', 'tokenizer.ggml.mask_token_id': '250001'}
2026-04-18 08:22:50.600 | Using fallback chat format: llama-2
2026-04-18 08:22:50.612 | INFO:     Started server process [8]
2026-04-18 08:22:50.612 | INFO:     Waiting for application startup.
2026-04-18 08:22:50.612 | INFO:     Application startup complete.
2026-04-18 08:22:50.613 | INFO:     Uvicorn running on http://localhost:5000 (Press CTRL+C to quit)
2026-04-18 08:24:53.546 | INFO:      - "POST /countIndexedDocuments HTTP/1.1" 200 OK
2026-04-18 08:31:43.821 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:31:54.885 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:31:59.293 | INFO:      - "POST /countIndexedDocuments HTTP/1.1" 200 OK
2026-04-18 08:32:08.087 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:32:24.895 | INFO:      - "POST /countIndexedDocuments HTTP/1.1" 200 OK
2026-04-18 08:32:30.165 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:32:40.720 | INFO:      - "POST /countIndexedDocuments HTTP/1.1" 200 OK
2026-04-18 08:32:45.157 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:32:52.472 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:33:05.367 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:33:13.744 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:33:30.049 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:33:50.027 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:34:04.498 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:34:18.333 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:34:32.213 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:34:57.256 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:35:17.997 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:35:45.119 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:35:50.476 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:04.408 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:15.969 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:29.781 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:34.240 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:35.958 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:43.205 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:49.570 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:36:52.528 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:37:15.112 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:37:49.587 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:07.645 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:15.071 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:21.868 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:34.122 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:51.848 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:38:55.004 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:10.146 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:12.968 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:21.193 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:30.104 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:35.949 | 2026-04-18T13:39:35+0000: [ERROR|doc_loader]: Error decoding source file (files__default: 12028)
2026-04-18 08:39:35.949 | Traceback (most recent call last):
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/doc_loader.py", line 130, in decode_source
2026-04-18 08:39:35.949 |     result = _loader_map[mimetype](source.file)
2026-04-18 08:39:35.949 |              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/doc_loader.py", line 52, in _load_docx
2026-04-18 08:39:35.949 |     return docx2txt.process(file).strip()
2026-04-18 08:39:35.949 |            ^^^^^^^^^^^^^^^^^^^^^^
2026-04-18 08:39:35.949 |   File "/usr/local/lib/python3.11/dist-packages/docx2txt/docx2txt.py", line 76, in process
2026-04-18 08:39:35.949 |     zipf = zipfile.ZipFile(docx)
2026-04-18 08:39:35.949 |            ^^^^^^^^^^^^^^^^^^^^^
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/zipfile.py", line 1329, in __init__
2026-04-18 08:39:35.949 |     self._RealGetContents()
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/zipfile.py", line 1396, in _RealGetContents
2026-04-18 08:39:35.949 |     raise BadZipFile("File is not a zip file")
2026-04-18 08:39:35.949 | zipfile.BadZipFile: File is not a zip file
2026-04-18 08:39:35.949 | Stack (most recent call last):
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/threading.py", line 1002, in _bootstrap
2026-04-18 08:39:35.949 |     self._bootstrap_inner()
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/threading.py", line 1045, in _bootstrap_inner
2026-04-18 08:39:35.949 |     self.run()
2026-04-18 08:39:35.949 |   File "/usr/local/lib/python3.11/dist-packages/anyio/_backends/_asyncio.py", line 1002, in run
2026-04-18 08:39:35.949 |     result = context.run(func, *args)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/controller.py", line 195, in wrapper
2026-04-18 08:39:35.949 |     return func(*args, **kwargs)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/controller.py", line 396, in _
2026-04-18 08:39:35.949 |     loaded_sources, not_added_sources = exec_in_proc(
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/utils.py", line 93, in exec_in_proc
2026-04-18 08:39:35.949 |     p.start()
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/process.py", line 121, in start
2026-04-18 08:39:35.949 |     self._popen = self._Popen(self)
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/context.py", line 224, in _Popen
2026-04-18 08:39:35.949 |     return _default_context.get_context().Process._Popen(process_obj)
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/context.py", line 281, in _Popen
2026-04-18 08:39:35.949 |     return Popen(process_obj)
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/popen_fork.py", line 19, in __init__
2026-04-18 08:39:35.949 |     self._launch(process_obj)
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/popen_fork.py", line 71, in _launch
2026-04-18 08:39:35.949 |     code = process_obj._bootstrap(parent_sentinel=child_r)
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
2026-04-18 08:39:35.949 |     self.run()
2026-04-18 08:39:35.949 |   File "/usr/lib/python3.11/multiprocessing/process.py", line 108, in run
2026-04-18 08:39:35.949 |     self._target(*self._args, **self._kwargs)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/utils.py", line 76, in exception_wrap
2026-04-18 08:39:35.949 |     resconn.send({ 'value': fun(*args, **kwargs), 'error': None })
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/injest.py", line 202, in embed_sources
2026-04-18 08:39:35.949 |     return _process_sources(vectordb, config, sources_filtered)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/injest.py", line 148, in _process_sources
2026-04-18 08:39:35.949 |     indocuments = _sources_to_indocuments(config, filtered_sources)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/injest.py", line 60, in _sources_to_indocuments
2026-04-18 08:39:35.949 |     content = decode_source(source)
2026-04-18 08:39:35.949 |   File "/app/context_chat_backend/chain/ingest/doc_loader.py", line 141, in decode_source
2026-04-18 08:39:35.949 |     logger.exception(f'Error decoding source file ({source.filename})', stack_info=True)
2026-04-18 08:39:35.960 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:38.869 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:48.836 | INFO:      - "PUT /loadSources HTTP/1.1" 200 OK
2026-04-18 08:39:56.641 |

The daemon config in nextcloud:

Deploy Daemon - Manual Install
Manual install daemon usually used for development. It cannot be set as default daemon.
ExApps installed: 1
Name: manual_install
Protocol: http
Host: 192.168.**:8780
ExApp direct communication (FRP disabled): false
Deploy config
Docker network: nextcloud-aio
Nextcloud URL: https:/something/index.php (removed name)
HaProxy password:
••••••••••••••
GPUs support: true
Compute device: CUDA (NVIDIA)

The HARP-API on the host runing nextcloud aio

services:
appapi-harp:
image: ghcr.io/nextcloud/nextcloud-appapi-harp:release
container_name: appapi-harp
hostname: appapi-harp
environment:
HP_SHARED_KEY: "..."
NC_INSTANCE_URL: "http://192.168....."
HP_EXAPPS_ADDRESS: "192.168.....0:8780"
HP_FRP_ADDRESS: "0.0.0.0:8782"
HP_TRUSTED_PROXY_IPS: "127.0.0.1,192.168.....0,192.168....."
volumes:
- /var/run/docker.sock:/var/run/docker.sock
- ./certs:/certs
network_mode: host

[kyteinsky]

thanks for the details.
it doesn't look like there is any issue with the setup, indexing also seems to be working fine, except a doc parsing error which might be due to a bad file but that file is just ignored and indexing continues.
pihole shouldn't really be involved since IPs are directly used in all places so no need for name resolution/DNS server.

what does the stats show?
occ context_chat:stats

in any case, if the process does seem to stop, I would suggest waiting for the next update which does significant changes to the indexing process and see if that improves things.
there are some fixes for files stuck in the indexing queue too.

[kyteinsky]

Deploy Daemon - Manual Install

any reason to use the manual daemon?
HaRP can be used with AIO if the docker network is set to nextcloud-aio, or there is the host network setup too.
also, app updates would be manual with the manual install daemon.

[knot-split]

context_chat:stats It doesn't report any errors, but what I've noticed is that it gets about to exactly the same place and then it just gets stuck. I've tried disabling and re-enabling the apps, and I seem to remember that it goes nicely until a spot where it just stops working, but it doesn't say that anything is wrong and for some reason also, when it gets stuck. It says that the number of files in the indexing queue is 113. . I'm sure I've seen that number twice at 2 different times when it got stuck. . I thought maybe it was the ad blocker, but I guess it's not.

When I 1st set up this container remotely, the harp container wasn't yet available on the AIO as a ready-made package. we are using the harp for this, , but I believe the problem with the remote set up is that if you don't choose a manual install then you can't stop nextcloud from trying to put the backend in the same container system. manual install seem to be the only way to get it on the remote computer.

[knot-split]

One other question to ask then:

If CCB is getting stuck on certain files, is there a way to tell it to ignore certain files?

[kyteinsky]

but I believe the problem with the remote set up is that if you don't choose a manual install then you can't stop nextcloud from trying to put the backend in the same container system. manual install seem to be the only way to get it on the remote computer.

this section might be what you're trying to achieve: https://docs.nextcloud.com/server/latest/admin_manual/exapps_management/DeployConfigurations.html#docker-on-a-remote-host-with-harp-container-on-the-local-host
it describes a system where Nextcloud and the HaRP container are on the main host but the ex-apps are installed on a remote host.

If CCB is getting stuck on certain files, is there a way to tell it to ignore certain files?

yeah the exact same figure of 113 might suggest this. We did a few improvements regarding stuck files recently so it would be nice if you could try the next version when it is released and see if the number goes down further.
if there are stuck files still, we can look at them case by case basis.

[knot-split]

this section might be what you're trying to achieve:

Yes, I was looking at those documentation files when I was trying to set things up. Without looking at it again just yet, I believe that those documents are not specifically targeting the AIO set up. If I can remember, there really wasn't any specific case wherein we had the AIO with a remote application running. But through trial and error over several weeks, I finally got something that would work by modifying those examples in the documentation.