Skip to content

🤖 Sync generated docs from sourcegraph/sourcegraph #2094

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sourcegraph-buildkite wants to merge 1 commit into
base: main
Choose a base branch
from
+443 −205
Open

🤖 Sync generated docs from sourcegraph/sourcegraph #2094

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions docs/admin/config/settings.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,9 @@ Settings options and their default values are shown below.
// Usually this setting is used in global and organization settings. If set in user settings, the message will only be displayed to that single user.
"notices": null,

// Whether to receive opted-in user notifications by email.
"notifications.email.enabled": true,

// Group of settings related to opening files in an editor.
"openInEditor": {
"custom.urlPattern": null,
Expand Down
6 changes: 3 additions & 3 deletions docs/ai/models.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,10 +7,10 @@ This reference lists the models currently used by [Sourcegraph AI features](/ai)

| Model | Used by |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| GPT-5.6 Luna | Change request file grouping, Deep Search compaction, [Code Finder](/api/mcp#code-finder), [Diff Tour](/diff-tour), Agentic Batch Changes workspace analysis, Agentic Batch Changes schema help |
| Claude Sonnet 5 | [Deep Search](/deep-search) |
| GPT-5.6 Luna | Change request file grouping, Deep Search compaction, [Diff Tour](/diff-tour), Agentic Batch Changes workspace analysis, Agentic Batch Changes schema help |
| Claude Sonnet 5.5 | Change request review report, [Deep Search](/deep-search) |
| GPT-6 Luna | Deep Search categorization, Deep Search answer summaries, Deep Search titles, Agentic Batch Changes titles, Agentic Batch Changes categorization, Agentic Batch Changes diagnostics, Coding agent low Finder, Coding agent medium Finder, Coding agent high Finder |
| GPT-5.4 Nano | [Smart Hover](/code-navigation/smart-hover) |
| Claude Haiku 5.5 | [Code Finder](/api/mcp#code-finder), [Smart Hover](/code-navigation/smart-hover) |
| Gemini 3.5 Flash-Lite | [Query Assist](/code-search/query-assist), Agentic Batch Changes repository search |
| Claude Opus 5.5 | [Agentic Batch Changes](/agentic-batch-changes), Agentic Batch Changes compaction |
| GPT-6.1 Sol | Coding agent low, Coding agent medium |
Expand Down
117 changes: 115 additions & 2 deletions docs/self-hosted/observability/alerts.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3225,6 +3225,43 @@ Generated query for critical alert: `max((max(src_codeintel_commit_graph_fast_pa

<br />

## worker: codeintel_uploads_writing_duration

<p class="subtitle">maximum time a processed SCIP upload has been waiting for the destination writer</p>

**Descriptions**

- <span class="badge badge-warning">warning</span> worker: 3600s+ maximum time a processed SCIP upload has been waiting for the destination writer for 15m0s
- <span class="badge badge-critical">critical</span> worker: 86400s+ maximum time a processed SCIP upload has been waiting for the destination writer for 15m0s

**Next steps**

- Check that the worker runs the codeintel-destination-writer job, and its logs for write or promotion errors.
- Check that the worker can reach the codeintel destination bundle store and the code intelligence database.
- Check the codeintel-db dashboard for resource saturation.
- More help interpreting this metric is available in the [dashboards reference](dashboards#worker-codeintel_uploads_writing_duration).
- **Silence this alert:** If you are aware of this alert and want to silence notifications for it, add the following to your site configuration and set a reminder to re-evaluate the alert:

```json
"observability.silenceAlerts": [
"warning_worker_codeintel_uploads_writing_duration",
"critical_worker_codeintel_uploads_writing_duration"
]
```

<sub>*Managed by the Sourcegraph Code Understanding team.*</sub>

<details>
<summary>Technical details</summary>

Generated query for warning alert: `max((max(src_codeintel_uploads_writing_max_age_seconds)) &gt;= 3600)`

Generated query for critical alert: `max((max(src_codeintel_uploads_writing_max_age_seconds)) &gt;= 86400)`

</details>

<br />

## worker: src_repoupdater_max_sync_backoff

<p class="subtitle">time since oldest sync</p>
Expand Down Expand Up @@ -7396,6 +7433,44 @@ Generated query for warning alert: `max((sum(increase(src_deepsearch_question_pr

<br />

## deepsearch: deepsearch_subagent_context_limit_rate

<p class="subtitle">subagent runs that hit the context limit by agent</p>

**Descriptions**

- <span class="badge badge-warning">warning</span> deepsearch: more than 10% of a subagent's runs hit the context limit over 1h
- <span class="badge badge-critical">critical</span> deepsearch: more than 25% of a subagent's runs hit the context limit over 1h

**Next steps**

- Subagents have no compaction. When the task plus accumulated tool results outgrow the input token budget of the subagent model, the subagent stops early. Usually it returns a summary of partial findings with a note that they may be incomplete (max_input_tokens_summarized); if even that summary does not fit or fails, the parent gets "sub-agent exceeded maximum input tokens; retry with a narrower task" instead (max_input_tokens). Either way the parent answers with incomplete findings or spends extra turns retrying.
- Check the "Agent tool calls" group for the same agent: a tool whose output grew recently, or a jump in its call rate, is the usual cause.
- A rise that starts at a deploy points at a changed subagent prompt, tool, or model input token budget.
- If the results of one tool dominate, tighten its result limits or the subagent instructions so it reads less per turn.
- More help interpreting this metric is available in the [dashboards reference](dashboards#deepsearch-deepsearch_subagent_context_limit_rate).
- **Silence this alert:** If you are aware of this alert and want to silence notifications for it, add the following to your site configuration and set a reminder to re-evaluate the alert:

```json
"observability.silenceAlerts": [
"warning_deepsearch_deepsearch_subagent_context_limit_rate",
"critical_deepsearch_deepsearch_subagent_context_limit_rate"
]
```

<sub>*Managed by the Sourcegraph Code Understanding team.*</sub>

<details>
<summary>Technical details</summary>

Custom query for warning alert: `max(((sum by (workload, agent) (increase(src_agent_runs_total{agent!="main",stop_reason=~"max_input_tokens\\|max_input_tokens_summarized",workload=~"deep_search"\}[1h])) / sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"deep_search"\}[1h])) * 100) and (sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"deep_search"}[1h])) &gt;= 10)) &gt; 10)`

Custom query for critical alert: `max(((sum by (workload, agent) (increase(src_agent_runs_total{agent!="main",stop_reason=~"max_input_tokens\\|max_input_tokens_summarized",workload=~"deep_search"\}[1h])) / sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"deep_search"\}[1h])) * 100) and (sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"deep_search"}[1h])) &gt;= 10)) &gt; 25)`

</details>

<br />

## deepsearch: deepsearch_response_truncated_no_answer

<p class="subtitle">truncated questions with no answer produced over 15m</p>
Expand Down Expand Up @@ -7505,6 +7580,44 @@ Generated query for critical alert: `max((max(src_workerutil_queue_duration_seco

<br />

## batch-change-agents: batch_change_agent_subagent_context_limit_rate

<p class="subtitle">subagent runs that hit the context limit by agent</p>

**Descriptions**

- <span class="badge badge-warning">warning</span> batch-change-agents: more than 10% of a subagent's runs hit the context limit over 1h
- <span class="badge badge-critical">critical</span> batch-change-agents: more than 25% of a subagent's runs hit the context limit over 1h

**Next steps**

- Subagents have no compaction. When the task plus accumulated tool results outgrow the input token budget of the subagent model, the subagent stops early. Usually it returns a summary of partial findings with a note that they may be incomplete (max_input_tokens_summarized); if even that summary does not fit or fails, the parent gets "sub-agent exceeded maximum input tokens; retry with a narrower task" instead (max_input_tokens). Either way the parent answers with incomplete findings or spends extra turns retrying.
- Check the "Agent tool calls" group for the same agent: a tool whose output grew recently, or a jump in its call rate, is the usual cause.
- A rise that starts at a deploy points at a changed subagent prompt, tool, or model input token budget.
- If the results of one tool dominate, tighten its result limits or the subagent instructions so it reads less per turn.
- More help interpreting this metric is available in the [dashboards reference](dashboards#batch-change-agents-batch_change_agent_subagent_context_limit_rate).
- **Silence this alert:** If you are aware of this alert and want to silence notifications for it, add the following to your site configuration and set a reminder to re-evaluate the alert:

```json
"observability.silenceAlerts": [
"warning_batch-change-agents_batch_change_agent_subagent_context_limit_rate",
"critical_batch-change-agents_batch_change_agent_subagent_context_limit_rate"
]
```

<sub>*Managed by the Sourcegraph Code Plane team.*</sub>

<details>
<summary>Technical details</summary>

Custom query for warning alert: `max(((sum by (workload, agent) (increase(src_agent_runs_total{agent!="main",stop_reason=~"max_input_tokens\\|max_input_tokens_summarized",workload=~"batch_change_agent"\}[1h])) / sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"batch_change_agent"\}[1h])) * 100) and (sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"batch_change_agent"}[1h])) &gt;= 10)) &gt; 10)`

Custom query for critical alert: `max(((sum by (workload, agent) (increase(src_agent_runs_total{agent!="main",stop_reason=~"max_input_tokens\\|max_input_tokens_summarized",workload=~"batch_change_agent"\}[1h])) / sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"batch_change_agent"\}[1h])) * 100) and (sum by (workload, agent) (increase(src_agent_runs_total\{agent!="main",workload=~"batch_change_agent"}[1h])) &gt;= 10)) &gt; 25)`

</details>

<br />

## batch-change-agents: batch_change_agent_runs_over_context_budget

<p class="subtitle">runs stopped because the conversation no longer fits the context budget over 1h</p>
Expand Down Expand Up @@ -7535,9 +7648,9 @@ Generated query for critical alert: `max((max(src_workerutil_queue_duration_seco
<details>
<summary>Technical details</summary>

Generated query for warning alert: `max((sum(increase(src_agent_runs_total{stop_reason="max_input_tokens",workload="batch_change_agent"}[1h]))) &gt; 0)`
Generated query for warning alert: `max((sum(increase(src_agent_runs_total{agent="main",stop_reason="max_input_tokens",workload="batch_change_agent"}[1h]))) &gt; 0)`

Generated query for critical alert: `max((sum(increase(src_agent_runs_total{stop_reason="max_input_tokens",workload="batch_change_agent"}[1h]))) &gt; 3)`
Generated query for critical alert: `max((sum(increase(src_agent_runs_total{agent="main",stop_reason="max_input_tokens",workload="batch_change_agent"}[1h]))) &gt; 3)`

</details>

Expand Down
Loading
Loading