Support for local ollama

[bitmorty]

Hello, I wanted to get started with sourcebot and trying to integrate my local instance of ollama. It works so far that I can select it in the input field. But every request shows "Error: No answer response was provided".

My model in config.json looks like this:

{
      "provider": "openai-compatible",
      "baseUrl": "http://host.docker.internal:11434",
      "model": "llama3.1:405b",
      "displayName": "ollama"
    }

Server logs:

latest: Pulling from sourcebot-dev/sourcebot
Digest: sha256:1e6b16404b37f150fc645981f479821fbf0cb9108d52c3a9e108d977a21372b1
Status: Image is up to date for ghcr.io/sourcebot-dev/sourcebot:latest
[Info] Sourcebot version: v4.6.4
[Warning] SOURCEBOT_ENCRYPTION_KEY is not set.
[Info] Loading environment variables from /data/.sourcebot/.secret
[Warning] AUTH_SECRET is not set.
[Info] Loading environment variables from /data/.sourcebot/.authjs-secret
[Warning] AUTH_URL is not set.
localhost:5432 - no response
[Info] Waiting for the database to be ready...
2025-08-13 07:51:21.321 UTC [14] LOG:  starting PostgreSQL 16.9 on aarch64-alpine-linux-musl, compiled by gcc (Alpine 13.2.1_git20231014) 13.2.1 20231014, 64-bit
2025-08-13 07:51:21.321 UTC [14] LOG:  listening on IPv6 address "::1", port 5432
2025-08-13 07:51:21.321 UTC [14] LOG:  listening on IPv4 address "127.0.0.1", port 5432
2025-08-13 07:51:21.323 UTC [14] LOG:  listening on Unix socket "/run/postgresql/.s.PGSQL.5432"
2025-08-13 07:51:21.327 UTC [19] LOG:  database system was interrupted; last known up at 2025-08-13 07:46:45 UTC
2025-08-13 07:51:21.704 UTC [19] LOG:  database system was not properly shut down; automatic recovery in progress
2025-08-13 07:51:21.708 UTC [19] LOG:  redo starts at 0/19E3B18
2025-08-13 07:51:21.710 UTC [19] LOG:  invalid record length at 0/19E8AE0: expected at least 24, got 0
2025-08-13 07:51:21.710 UTC [19] LOG:  redo done at 0/19E8AA8 system usage: CPU: user: 0.00 s, system: 0.00 s, elapsed: 0.00 s
2025-08-13 07:51:21.715 UTC [17] LOG:  checkpoint starting: end-of-recovery immediate wait
2025-08-13 07:51:21.719 UTC [17] LOG:  checkpoint complete: wrote 11 buffers (0.1%); 0 WAL file(s) added, 0 removed, 0 recycled; write=0.002 s, sync=0.001 s, total=0.005 s; sync files=11, longest=0.001 s, average=0.001 s; distance=20 kB, estimate=20 kB; lsn=0/19E8AE0, redo lsn=0/19E8AE0
2025-08-13 07:51:21.720 UTC [14] LOG:  database system is ready to accept connections
localhost:5432 - accepting connections
Database 'sourcebot' already exists; skipping creation.
[Info] Running database migration...
Prisma schema loaded from prisma/schema.prisma
Datasource "db": PostgreSQL database "sourcebot", schema "public" at "localhost:5432"

34 migrations found in prisma/migrations


No pending migrations to apply.
2025-08-13 07:51:23,262 CRIT Supervisor is running as root.  Privileges were not dropped because no user is specified in the config file.  If you intend to run as root, you can set user=root in the config file to avoid this message.
2025-08-13 07:51:23,263 INFO supervisord started with pid 7
2025-08-13 07:51:24,271 INFO spawned: 'backend' with pid 156
2025-08-13 07:51:24,274 INFO spawned: 'redis' with pid 157
2025-08-13 07:51:24,277 INFO spawned: 'web' with pid 160
2025-08-13 07:51:24,279 INFO spawned: 'zoekt' with pid 161
157:C 13 Aug 2025 07:51:24.282 * oO0OoO0OoO0Oo Redis is starting oO0OoO0OoO0Oo
157:C 13 Aug 2025 07:51:24.282 * Redis version=7.2.9, bits=64, commit=00000000, modified=0, pid=157, just started
157:C 13 Aug 2025 07:51:24.282 * Configuration loaded
157:M 13 Aug 2025 07:51:24.282 * monotonic clock: POSIX clock_gettime
157:M 13 Aug 2025 07:51:24.283 * Running mode=standalone, port=6379.
157:M 13 Aug 2025 07:51:24.283 * Server initialized
157:M 13 Aug 2025 07:51:24.284 * Loading RDB produced by version 7.2.9
157:M 13 Aug 2025 07:51:24.284 * RDB age 5 seconds
157:M 13 Aug 2025 07:51:24.284 * RDB memory usage when created 1.75 Mb
157:M 13 Aug 2025 07:51:24.284 * Done loading RDB, keys loaded: 13, keys expired: 0.
157:M 13 Aug 2025 07:51:24.285 * DB loaded from disk: 0.001 seconds
157:M 13 Aug 2025 07:51:24.285 * Ready to accept connections tcp
[zoekt] | {"SeverityText":"WARN","Timestamp":1755071484315680004,"InstrumentationScope":"metricsRegistration.mountPointInfo.deviceNameDiscovery","Caller":"[email protected]/info.go:48","Function":"github.com/sourcegraph/mountinfo.NewCollector","Body":"skipping metric registration","Resource":{"service.name":"zoekt-webserver","service.instance.id":"0.0.0.0"},"Attributes":{"mountName":"indexDir","mountFilePath":"/data/.sourcebot/index","reason":"failed to discover device name","error":"discovering device path: discoverSysfsDevicePath: failed to evaluate sysfs symlink \"/sys/dev/block/0:44\": lstat /sys/dev/block/0:44: no such file or directory"}}
[zoekt] | 2025/08/13 07:51:24 [INFO] loading 1 shard(s): 1_1_v16.00000.zoekt
[web] |   ▲ Next.js 14.2.30
[web] |   - Local:        http://localhost:3000
[web] |   - Network:      http://0.0.0.0:3000
[web] |
[web] |  ✓ Starting...
[web] |  ✓ Ready in 130ms
[web] | 2025-08-13T07:51:24.636Z info: [web-initialize] Upserted connection with name 'my-app-connection'. Connection ID: 34
[backend] | 2025-08-13T07:51:24.688Z info: [repo-manager] Validating indexed repos have shards...
[backend] | 2025-08-13T07:51:24.690Z info: [prometheus-client] Prometheus metrics server is running on port 3060
[backend] | 2025-08-13T07:51:24.692Z info: [backend-main] Connected to redis
[backend] | 2025-08-13T07:51:24.718Z info: [repo-manager] Found 1 repos in the DB marked as INDEXED
[backend] | 2025-08-13T07:51:24.719Z info: [repo-manager] Done validating indexed repos have shards
[backend] | 2025-08-13T07:51:25.702Z info: [connection-manager] Added job to queue for connection my-app-connection (id: 34)
2025-08-13 07:51:25,703 INFO success: backend entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
2025-08-13 07:51:25,703 INFO success: redis entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
2025-08-13 07:51:25,703 INFO success: web entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
2025-08-13 07:51:25,703 INFO success: zoekt entered RUNNING state, process has stayed up for > than 1 seconds (startsecs)
[backend] | 2025-08-13T07:51:26.001Z info: [github] Fetching repository info for company/my-app...
[backend] | 2025-08-13T07:51:26.940Z info: [github] Found info for repository company/my-app in 939ms
[backend] | 2025-08-13T07:51:26.948Z info: [connection-manager] Deleted all RepoToConnection records for connection my-app-connection (id: 34) in 2.4179159999998774ms
[backend] | 2025-08-13T07:51:26.954Z info: [connection-manager] Upserted repo company/my-app (id: 232752691) in 5.81795800000009ms
[backend] | 2025-08-13T07:51:26.954Z info: [connection-manager] Upserted 1 repos for connection my-app-connection (id: 34) in 6.101250000000164ms
[backend] | 2025-08-13T07:51:26.957Z info: [connection-manager] Connection sync job for connection my-app-connection (id: 34, jobId: 27) completed
[web] | [edge] Sentry was not initialized
[web] | 2025-08-13T07:51:51.962Z error: [web-actions] Not Found
[web] | AI_APICallError: Not Found
[web] |     at /app/packages/web/.next/server/chunks/4348.js:53:9552
[web] |     at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
[web] |     at async q (/app/packages/web/.next/server/chunks/4348.js:53:7691)
[web] |     at async c.doGenerate (/app/packages/web/.next/server/chunks/4348.js:29:92183)
[web] |     at async fn (/app/packages/web/.next/server/chunks/4348.js:94:35148)
[web] |     at async /app/packages/web/.next/server/chunks/4348.js:94:25055
[web] |     at async tz (/app/packages/web/.next/server/chunks/4348.js:94:26761)
[web] |     at async fn (/app/packages/web/.next/server/chunks/4348.js:94:34310)
[web] |     at async /app/packages/web/.next/server/chunks/4348.js:94:25055
[web] |     at async t0 (/app/packages/web/.next/server/chunks/4348.js:94:33520)
[web] | 2025-08-13T07:51:51.967Z error: [chat-agent] [object Object]
[web] | 2025-08-13T07:51:51.967Z error: [chat-api] Not Found
[web] | Error: Not Found
[web] |     at /app/packages/web/.next/server/chunks/4348.js:96:10874
[web] |     at runUpdateMessageJob (/app/packages/web/.next/server/chunks/4348.js:96:11267)
[web] |     at Object.transform (/app/packages/web/.next/server/chunks/4348.js:96:4726)
[web] |     at invokePromiseCallback (node:internal/webstreams/util:181:10)
[web] |     at Object.transformAlgorithm (node:internal/webstreams/util:186:23)
[web] |     at transformStreamDefaultControllerPerformTransform (node:internal/webstreams/transformstream:519:37)
[web] |     at transformStreamDefaultSinkWriteAlgorithm (node:internal/webstreams/transformstream:565:10)
[web] |     at node:internal/webstreams/transformstream:370:16
[web] |     at writableStreamDefaultControllerProcessWrite (node:internal/webstreams/writablestream:1120:5)
[web] |     at writableStreamDefaultControllerAdvanceQueueIfNeeded (node:internal/webstreams/writablestream:1235:5)

[brendan-kellam]

Are there any relevant server side logs?

[bitmorty]

Are there any relevant server side logs?

Sorry. I updated my post with the logs. At the end you can see what happens when the frontend shows me the error.