kql-panopticon 0.3.0

KQL tooling for Azure Log Analytics - concurrent multi-workspace queries, chained investigations, HTTP enrichment, and automated reports
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
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
use crate::client::{Client, QueryResponse, Table};
use crate::error::{KqlPanopticonError, Result};
use crate::workspace::Workspace;
use chrono::{DateTime, Local};
use log::{debug, info, warn};
use std::path::{Path, PathBuf};
use std::time::{Duration, Instant};
use tokio::fs;
use tokio::io::AsyncWriteExt;

/// Generate a unique temp file path to avoid collisions during concurrent executions
fn generate_unique_temp_path(base_path: &Path, extension: &str) -> PathBuf {
    let timestamp = chrono::Local::now().format("%Y%m%d_%H%M%S_%3f");
    let pid = std::process::id();

    // Create unique temp filename: base.tmp.{timestamp}_{pid}.{extension}
    let mut temp_path = base_path.to_path_buf();
    let temp_filename = format!(
        "{}.tmp.{}_{}.{}",
        base_path.file_stem().unwrap_or_default().to_string_lossy(),
        timestamp,
        pid,
        extension
    );
    temp_path.set_file_name(temp_filename);
    temp_path
}

/// Settings for query execution
#[derive(Debug, Clone, serde::Serialize, serde::Deserialize)]
#[serde(default)]
pub struct QuerySettings {
    /// Base output folder for all results
    pub output_folder: PathBuf,

    /// Job name used for file naming
    pub job_name: String,

    /// Export results as CSV files
    pub export_csv: bool,

    /// Export results as JSON files
    pub export_json: bool,

    /// Parse nested dynamic fields into JSON objects (only affects JSON export)
    pub parse_dynamics: bool,
}

impl Default for QuerySettings {
    fn default() -> Self {
        Self {
            output_folder: PathBuf::from("./output"),
            job_name: "query".to_string(),
            export_csv: true,
            export_json: false,
            parse_dynamics: true,
        }
    }
}

impl QuerySettings {
    #[allow(dead_code)]
    pub fn new(output_folder: impl Into<PathBuf>, job_name: impl Into<String>) -> Self {
        Self {
            output_folder: output_folder.into(),
            job_name: job_name.into(),
            export_csv: true,
            export_json: false,
            parse_dynamics: true,
        }
    }

    pub fn with_formats(
        output_folder: impl Into<PathBuf>,
        job_name: impl Into<String>,
        export_csv: bool,
        export_json: bool,
        parse_dynamics: bool,
    ) -> Self {
        Self {
            output_folder: output_folder.into(),
            job_name: job_name.into(),
            export_csv,
            export_json,
            parse_dynamics,
        }
    }
}

/// Result of a single query job execution
#[derive(Debug, Clone)]
pub struct QueryJobResult {
    /// Workspace that was queried
    pub workspace_id: String,

    /// Workspace name
    pub workspace_name: String,

    /// Query that was executed
    pub query: String,

    /// Execution result
    pub result: Result<JobSuccess>,

    /// Time taken to execute
    pub elapsed: Duration,

    /// Timestamp when the job completed
    pub timestamp: DateTime<Local>,
}

/// Success information for a completed job
#[derive(Debug, Clone, serde::Serialize, serde::Deserialize)]
pub struct JobSuccess {
    /// Number of rows returned
    pub row_count: usize,

    /// Number of pages fetched (for paginated queries)
    #[allow(dead_code)]
    pub page_count: usize,

    /// Output file path
    pub output_path: PathBuf,

    /// File size in bytes
    pub file_size: u64,
}

/// Individual query job
struct QueryJob {
    workspace: Workspace,
    query: String,
    settings: QuerySettings,
    timestamp: String,
}

/// Helper for streaming CSV writes to a temporary file
struct StreamingCsvWriter {
    temp_path: PathBuf,
    file: tokio::fs::File,
    row_count: usize,
    page_count: usize,
    buffer: Vec<String>,
    buffer_size: usize,
}

impl StreamingCsvWriter {
    /// Create a new streaming CSV writer
    async fn new(temp_path: PathBuf, buffer_size: usize) -> Result<Self> {
        let file = tokio::fs::File::create(&temp_path).await?;
        Ok(Self {
            temp_path,
            file,
            row_count: 0,
            page_count: 0,
            buffer: Vec::with_capacity(buffer_size),
            buffer_size,
        })
    }

    /// Write CSV header
    async fn write_header(&mut self, table: &Table) -> Result<()> {
        let headers: Vec<String> = table.columns.iter().map(|col| col.name.clone()).collect();
        let header_line = format!("{}\n", headers.join(","));
        self.file.write_all(header_line.as_bytes()).await?;
        Ok(())
    }

    /// Add rows from a page to the buffer
    fn add_page(&mut self, table: &Table, format_fn: &impl Fn(&serde_json::Value) -> String) {
        self.page_count += 1;
        for row in &table.rows {
            if let Some(row_array) = row.as_array() {
                let row_strings: Vec<String> = row_array.iter().map(format_fn).collect();
                self.buffer.push(format!("{}\n", row_strings.join(",")));
                self.row_count += 1;
            }
        }
    }

    /// Flush buffer to disk if it exceeds buffer_size
    async fn flush_if_needed(&mut self) -> Result<()> {
        if self.buffer.len() >= self.buffer_size {
            self.flush().await?;
        }
        Ok(())
    }

    /// Flush buffer to disk
    async fn flush(&mut self) -> Result<()> {
        if !self.buffer.is_empty() {
            let content = self.buffer.join("");
            self.file.write_all(content.as_bytes()).await?;
            self.buffer.clear();
        }
        Ok(())
    }

    /// Finalize the file and move to final location
    async fn finalize(mut self, final_path: &Path) -> Result<usize> {
        // Flush any remaining buffered data
        self.flush().await?;

        // Ensure all data is written to disk
        self.file.sync_all().await?;

        // Close the file
        drop(self.file);

        // Move temp file to final location
        tokio::fs::rename(&self.temp_path, final_path).await?;

        Ok(self.row_count)
    }

    /// Clean up temp file on error
    async fn cleanup(self) -> Result<()> {
        drop(self.file);
        if self.temp_path.exists() {
            tokio::fs::remove_file(&self.temp_path).await?;
        }
        Ok(())
    }

    /// Save partial results when pagination fails
    async fn save_partial(mut self, output_path: &Path) -> Result<(usize, PathBuf)> {
        // Flush any remaining buffered data
        self.flush().await?;

        // Ensure all data is written to disk
        self.file.sync_all().await?;

        // Close the file
        drop(self.file);

        // Create partial result filename
        let partial_path = output_path.with_extension("partial.csv");

        // Move temp file to partial location
        tokio::fs::rename(&self.temp_path, &partial_path).await?;

        warn!(
            "Saved partial results ({} rows, {} pages) to: {}",
            self.row_count,
            self.page_count,
            partial_path.display()
        );

        Ok((self.row_count, partial_path))
    }
}

/// Helper for streaming JSON writes to a temporary file
struct StreamingJsonWriter {
    temp_path: PathBuf,
    file: tokio::fs::File,
    row_count: usize,
    page_count: usize,
    buffer: Vec<serde_json::Value>,
    buffer_size: usize,
    table_columns: Option<Vec<crate::client::Column>>,
    parse_dynamics: bool,
}

impl StreamingJsonWriter {
    /// Create a new streaming JSON writer
    async fn new(temp_path: PathBuf, buffer_size: usize, parse_dynamics: bool) -> Result<Self> {
        let file = tokio::fs::File::create(&temp_path).await?;
        Ok(Self {
            temp_path,
            file,
            row_count: 0,
            page_count: 0,
            buffer: Vec::with_capacity(buffer_size),
            buffer_size,
            table_columns: None,
            parse_dynamics,
        })
    }

    /// Set table columns (must be called before adding pages)
    fn set_columns(&mut self, columns: Vec<crate::client::Column>) {
        self.table_columns = Some(columns);
    }

    /// Add rows from a page to the buffer
    fn add_page(&mut self, table: &Table) -> Result<()> {
        if self.table_columns.is_none() {
            return Err(KqlPanopticonError::InvalidConfiguration(
                "Table columns not set before adding page".to_string(),
            ));
        }

        self.page_count += 1;
        let columns = self.table_columns.as_ref().unwrap();

        for row in &table.rows {
            if let Some(row_array) = row.as_array() {
                let mut row_object = serde_json::Map::new();
                for (idx, value) in row_array.iter().enumerate() {
                    if let Some(column) = columns.get(idx) {
                        let processed_value =
                            if self.parse_dynamics && column.column_type == "dynamic" {
                                Self::parse_dynamic_value(value)
                            } else {
                                value.clone()
                            };
                        row_object.insert(column.name.clone(), processed_value);
                    }
                }
                self.buffer.push(serde_json::Value::Object(row_object));
                self.row_count += 1;
            }
        }

        Ok(())
    }

    /// Flush buffer to disk if it exceeds buffer_size
    async fn flush_if_needed(&mut self) -> Result<()> {
        if self.buffer.len() >= self.buffer_size {
            self.flush().await?;
        }
        Ok(())
    }

    /// Flush buffer to disk (as newline-delimited JSON)
    async fn flush(&mut self) -> Result<()> {
        if !self.buffer.is_empty() {
            for value in &self.buffer {
                let line = serde_json::to_string(value)?;
                self.file.write_all(line.as_bytes()).await?;
                self.file.write_all(b"\n").await?;
            }
            self.buffer.clear();
        }
        Ok(())
    }

    /// Finalize the file and move to final location with metadata
    async fn finalize(
        mut self,
        final_path: &Path,
        workspace: &Workspace,
        timestamp: &str,
        query: &str,
    ) -> Result<usize> {
        // Flush any remaining buffered data
        self.flush().await?;

        // Close the temp file
        drop(self.file);

        // Read all rows from temp file
        let temp_content = tokio::fs::read_to_string(&self.temp_path).await?;
        let rows: Vec<serde_json::Value> = temp_content
            .lines()
            .filter(|line| !line.is_empty())
            .map(|line| serde_json::from_str(line).unwrap_or(serde_json::Value::Null))
            .collect();

        // Build final JSON with metadata
        let columns = self.table_columns.as_ref().ok_or_else(|| {
            KqlPanopticonError::InvalidConfiguration("Table columns not set".to_string())
        })?;

        let output = serde_json::json!({
            "metadata": {
                "workspace": workspace.name,
                "workspace_id": workspace.workspace_id,
                "subscription": workspace.subscription_name,
                "timestamp": timestamp,
                "query": query,
                "row_count": self.row_count,
                "page_count": self.page_count,
            },
            "columns": columns.iter().map(|col| {
                serde_json::json!({
                    "name": col.name,
                    "type": col.column_type,
                })
            }).collect::<Vec<_>>(),
            "rows": rows,
        });

        // Write final JSON to destination
        let json_content = serde_json::to_string_pretty(&output)?;
        tokio::fs::write(final_path, json_content).await?;

        // Clean up temp file
        tokio::fs::remove_file(&self.temp_path).await?;

        Ok(self.row_count)
    }

    /// Clean up temp file on error
    async fn cleanup(self) -> Result<()> {
        drop(self.file);
        if self.temp_path.exists() {
            tokio::fs::remove_file(&self.temp_path).await?;
        }
        Ok(())
    }

    /// Save partial results when pagination fails
    async fn save_partial(
        mut self,
        output_path: &Path,
        workspace: &Workspace,
        timestamp: &str,
        query: &str,
    ) -> Result<(usize, PathBuf)> {
        // Flush any remaining buffered data
        self.flush().await?;

        // Close the temp file
        drop(self.file);

        // Read all rows from temp file
        let temp_content = tokio::fs::read_to_string(&self.temp_path).await?;
        let rows: Vec<serde_json::Value> = temp_content
            .lines()
            .filter(|line| !line.is_empty())
            .map(|line| serde_json::from_str(line).unwrap_or(serde_json::Value::Null))
            .collect();

        // Build partial JSON output with metadata indicating incompleteness
        let output = serde_json::json!({
            "workspace": {
                "name": workspace.name,
                "id": workspace.workspace_id,
                "resource_group": workspace.resource_group,
                "subscription_id": workspace.subscription_id,
                "subscription_name": workspace.subscription_name,
            },
            "query": query,
            "timestamp": timestamp,
            "partial": true,
            "rows_retrieved": self.row_count,
            "pages_retrieved": self.page_count,
            "note": "This result is incomplete due to pagination failure. Only partial data is available.",
            "rows": rows,
        });

        // Create partial result filename
        let partial_path = output_path.with_extension("partial.json");

        // Write partial JSON to destination
        let json_content = serde_json::to_string_pretty(&output)?;
        tokio::fs::write(&partial_path, json_content).await?;

        // Clean up temp file
        tokio::fs::remove_file(&self.temp_path).await?;

        warn!(
            "Saved partial results ({} rows, {} pages) to: {}",
            self.row_count,
            self.page_count,
            partial_path.display()
        );

        Ok((self.row_count, partial_path))
    }

    /// Recursively parse dynamic values that might be JSON strings
    fn parse_dynamic_value(value: &serde_json::Value) -> serde_json::Value {
        match value {
            serde_json::Value::String(s) => {
                if let Ok(parsed) = serde_json::from_str::<serde_json::Value>(s) {
                    Self::parse_dynamic_value(&parsed)
                } else {
                    value.clone()
                }
            }
            serde_json::Value::Array(arr) => {
                let processed: Vec<_> = arr.iter().map(Self::parse_dynamic_value).collect();
                serde_json::Value::Array(processed)
            }
            serde_json::Value::Object(obj) => {
                let mut processed = serde_json::Map::new();
                for (k, v) in obj {
                    processed.insert(k.clone(), Self::parse_dynamic_value(v));
                }
                serde_json::Value::Object(processed)
            }
            _ => value.clone(),
        }
    }
}

/// Builder for creating and executing query jobs
pub struct QueryJobBuilder {
    workspaces: Vec<Workspace>,
    queries: Vec<String>,
    settings: Option<QuerySettings>,
}

impl QueryJobBuilder {
    /// Create a new query job builder
    pub fn new() -> Self {
        Self {
            workspaces: Vec::new(),
            queries: Vec::new(),
            settings: None,
        }
    }

    /// Add workspaces to query
    pub fn workspaces(mut self, workspaces: Vec<Workspace>) -> Self {
        self.workspaces = workspaces;
        self
    }

    /// Add queries to execute
    pub fn queries(mut self, queries: Vec<String>) -> Self {
        self.queries = queries;
        self
    }

    /// Set query settings
    pub fn settings(mut self, settings: QuerySettings) -> Self {
        self.settings = Some(settings);
        self
    }

    /// Generate timestamp string in format: YYYY-MM-DD_HH-MM-SS
    fn generate_timestamp() -> String {
        let now: DateTime<Local> = Local::now();
        now.format("%Y-%m-%d_%H-%M-%S").to_string()
    }

    /// Execute all query jobs
    pub async fn execute(self, client: &Client) -> Result<Vec<QueryJobResult>> {
        let settings = self.settings.ok_or_else(|| {
            KqlPanopticonError::InvalidConfiguration("QuerySettings not provided".to_string())
        })?;

        if self.workspaces.is_empty() {
            return Err(KqlPanopticonError::InvalidConfiguration(
                "No workspaces provided".to_string(),
            ));
        }

        if self.queries.is_empty() {
            return Err(KqlPanopticonError::InvalidConfiguration(
                "No queries provided".to_string(),
            ));
        }

        let timestamp = Self::generate_timestamp();

        // Create all jobs (cartesian product of workspaces � queries)
        let mut jobs = Vec::new();
        for workspace in self.workspaces {
            for query in &self.queries {
                jobs.push(QueryJob {
                    workspace: workspace.clone(),
                    query: query.clone(),
                    settings: settings.clone(),
                    timestamp: timestamp.clone(),
                });
            }
        }

        info!("Executing {} query job(s)", jobs.len());

        // Execute all jobs concurrently
        let mut tasks = Vec::new();
        for job in jobs {
            let client = client.clone();
            let task = tokio::spawn(async move { job.execute(&client).await });
            tasks.push(task);
        }

        // Collect results
        let mut results = Vec::new();
        for task in tasks {
            match task.await {
                Ok(result) => results.push(result),
                Err(e) => {
                    warn!("Task panicked: {}", e);
                }
            }
        }

        Ok(results)
    }
}

impl Default for QueryJobBuilder {
    fn default() -> Self {
        Self::new()
    }
}

impl QueryJob {
    /// Execute this query job
    async fn execute(self, client: &Client) -> QueryJobResult {
        let start = Instant::now();

        debug!(
            "Executing query on workspace '{}' ({})",
            self.workspace.name, self.workspace.workspace_id
        );

        let result = self.execute_and_save(client).await;
        let elapsed = start.elapsed();

        match &result {
            Ok(success) => {
                info!(
                    " Completed: {} rows written to {} ({:.2}s)",
                    success.row_count,
                    success.output_path.display(),
                    elapsed.as_secs_f64()
                );
            }
            Err(e) => {
                warn!(
                    " Failed on workspace '{}': {} ({:.2}s)",
                    self.workspace.name,
                    e,
                    elapsed.as_secs_f64()
                );
            }
        }

        QueryJobResult {
            workspace_id: self.workspace.workspace_id.clone(),
            workspace_name: self.workspace.name.clone(),
            query: self.query.clone(),
            result,
            elapsed,
            timestamp: Local::now(),
        }
    }

    /// Execute query and save to configured formats (CSV and/or JSON) with pagination support
    async fn execute_and_save(&self, client: &Client) -> Result<JobSuccess> {
        // Build output directory: output_folder/subscription_name/workspace_name/timestamp/
        let normalized_subscription = Workspace::normalize_name(&self.workspace.subscription_name);
        let normalized_workspace = Workspace::normalize_name(&self.workspace.name);

        let output_dir = self
            .settings
            .output_folder
            .join(normalized_subscription)
            .join(normalized_workspace)
            .join(&self.timestamp);

        // Create directory structure
        fs::create_dir_all(&output_dir).await?;

        let mut row_count = 0;
        let mut page_count = 0;
        let mut total_file_size = 0u64;
        let mut primary_output_path = None;

        // Export as CSV if enabled
        if self.settings.export_csv {
            let csv_path = output_dir.join(format!("{}.csv", self.settings.job_name));
            let (rows, pages) = self.write_csv_streaming(client, &csv_path).await?;
            row_count = rows;
            page_count = pages;
            let metadata = fs::metadata(&csv_path).await?;
            total_file_size += metadata.len();
            if primary_output_path.is_none() {
                primary_output_path = Some(csv_path);
            }
        }

        // Export as JSON if enabled
        if self.settings.export_json {
            let json_path = output_dir.join(format!("{}.json", self.settings.job_name));
            let (rows, pages) = self.write_json_streaming(client, &json_path).await?;
            row_count = rows;
            page_count = pages;
            let metadata = fs::metadata(&json_path).await?;
            total_file_size += metadata.len();
            if primary_output_path.is_none() {
                primary_output_path = Some(json_path);
            }
        }

        let output_path = primary_output_path.ok_or_else(|| {
            KqlPanopticonError::InvalidConfiguration(
                "No export format enabled (CSV or JSON required)".to_string(),
            )
        })?;

        Ok(JobSuccess {
            row_count,
            page_count,
            output_path,
            file_size: total_file_size,
        })
    }

    /// Write query response to CSV file with streaming and pagination
    async fn write_csv_streaming(
        &self,
        client: &Client,
        output_path: &Path,
    ) -> Result<(usize, usize)> {
        // Create unique temp file path to avoid collisions during concurrent executions
        let temp_path = generate_unique_temp_path(output_path, "csv");

        // Buffer 100 pages before flushing to disk (adjustable)
        const PAGE_BUFFER_SIZE: usize = 100;

        let mut writer = StreamingCsvWriter::new(temp_path.clone(), PAGE_BUFFER_SIZE).await?;

        // Execute first query with retry logic
        let timeout = client.query_timeout();
        let retry_count = client.retry_count();
        let mut response = self
            .execute_with_retry(client, timeout, retry_count)
            .await?;

        if response.tables.is_empty() {
            writer.cleanup().await?;
            return Err(KqlPanopticonError::QueryExecutionFailed(
                "Query returned no tables".to_string(),
            ));
        }

        // Write header from first table
        let table = &response.tables[0];
        writer.write_header(table).await?;

        // Process first page
        writer.add_page(table, &|value| self.format_csv_value(value));
        writer.flush_if_needed().await?;

        // Follow pagination links
        while let Some(ref next_link) = response.next_link {
            debug!("Fetching next page: {} rows so far", writer.row_count);

            let page_future = client.query_next_page(next_link);
            response =
                match tokio::time::timeout(timeout, page_future).await {
                    Ok(Ok(page)) => page,
                    Ok(Err(e)) => {
                        // Pagination failed, save partial results
                        let (rows, partial_path) = writer.save_partial(output_path).await?;
                        return Err(KqlPanopticonError::QueryExecutionFailed(format!(
                            "Pagination failed after {} rows (saved to {}): {}",
                            rows,
                            partial_path.display(),
                            e
                        )));
                    }
                    Err(_) => {
                        // Timeout, save partial results
                        let (rows, partial_path) = writer.save_partial(output_path).await?;
                        return Err(KqlPanopticonError::QueryExecutionFailed(format!(
                        "Pagination timed out after {} seconds, {} rows retrieved (saved to {})",
                        timeout.as_secs(), rows, partial_path.display()
                    )));
                    }
                };

            if !response.tables.is_empty() {
                let table = &response.tables[0];
                writer.add_page(table, &|value| self.format_csv_value(value));
                writer.flush_if_needed().await?;
            }
        }

        // Finalize: flush remaining buffer and move to final location
        let row_count = writer.row_count;
        let page_count = writer.page_count;

        match writer.finalize(output_path).await {
            Ok(_) => Ok((row_count, page_count)),
            Err(e) => {
                // Try to cleanup temp file on finalization error
                let _ = tokio::fs::remove_file(&temp_path).await;
                Err(e)
            }
        }
    }

    /// Write query response to JSON file with streaming and pagination
    async fn write_json_streaming(
        &self,
        client: &Client,
        output_path: &Path,
    ) -> Result<(usize, usize)> {
        // Create unique temp file path to avoid collisions during concurrent executions
        let temp_path = generate_unique_temp_path(output_path, "json");

        // Buffer 100 pages before flushing to disk (adjustable)
        const PAGE_BUFFER_SIZE: usize = 100;

        let mut writer = StreamingJsonWriter::new(
            temp_path.clone(),
            PAGE_BUFFER_SIZE,
            self.settings.parse_dynamics,
        )
        .await?;

        // Execute first query with retry logic
        let timeout = client.query_timeout();
        let retry_count = client.retry_count();
        let mut response = self
            .execute_with_retry(client, timeout, retry_count)
            .await?;

        if response.tables.is_empty() {
            writer.cleanup().await?;
            return Err(KqlPanopticonError::QueryExecutionFailed(
                "Query returned no tables".to_string(),
            ));
        }

        // Set columns from first table
        let table = &response.tables[0];
        writer.set_columns(table.columns.clone());

        // Process first page
        writer.add_page(table)?;
        writer.flush_if_needed().await?;

        // Follow pagination links
        while let Some(ref next_link) = response.next_link {
            debug!("Fetching next page: {} rows so far", writer.row_count);

            let page_future = client.query_next_page(next_link);
            response = match tokio::time::timeout(timeout, page_future).await {
                Ok(Ok(page)) => page,
                Ok(Err(e)) => {
                    // Pagination failed, save partial results
                    let (rows, partial_path) = writer
                        .save_partial(output_path, &self.workspace, &self.timestamp, &self.query)
                        .await?;
                    return Err(KqlPanopticonError::QueryExecutionFailed(format!(
                        "Pagination failed after {} rows (saved to {}): {}",
                        rows,
                        partial_path.display(),
                        e
                    )));
                }
                Err(_) => {
                    // Timeout, save partial results
                    let (rows, partial_path) = writer
                        .save_partial(output_path, &self.workspace, &self.timestamp, &self.query)
                        .await?;
                    return Err(KqlPanopticonError::QueryExecutionFailed(format!(
                        "Pagination timed out after {} seconds, {} rows retrieved (saved to {})",
                        timeout.as_secs(),
                        rows,
                        partial_path.display()
                    )));
                }
            };

            if !response.tables.is_empty() {
                let table = &response.tables[0];
                writer.add_page(table)?;
                writer.flush_if_needed().await?;
            }
        }

        // Finalize: flush remaining buffer, wrap with metadata, and move to final location
        let row_count = writer.row_count;
        let page_count = writer.page_count;

        match writer
            .finalize(output_path, &self.workspace, &self.timestamp, &self.query)
            .await
        {
            Ok(_) => Ok((row_count, page_count)),
            Err(e) => {
                // Try to cleanup temp file on finalization error
                let _ = tokio::fs::remove_file(&temp_path).await;
                Err(e)
            }
        }
    }

    /// Execute query with retry logic and timeout
    async fn execute_with_retry(
        &self,
        client: &Client,
        timeout: Duration,
        retry_count: u32,
    ) -> Result<QueryResponse> {
        let mut last_error = None;
        let max_attempts = retry_count + 1; // retry_count of 0 means 1 attempt total

        for attempt in 0..max_attempts {
            if attempt > 0 {
                // Determine backoff duration based on last error
                let backoff = match &last_error {
                    Some(KqlPanopticonError::RateLimitExceeded { retry_after }) => {
                        // Use Azure's specified retry-after time
                        info!(
                            "Rate limited on workspace '{}'. Waiting {} seconds before retry (attempt {}/{})",
                            self.workspace.name,
                            retry_after,
                            attempt + 1,
                            max_attempts
                        );
                        Duration::from_secs(*retry_after)
                    }
                    _ => {
                        // Standard exponential backoff: 1s, 2s, 4s, 8s, etc.
                        debug!(
                            "Retrying query on workspace '{}' (attempt {}/{})",
                            self.workspace.name,
                            attempt + 1,
                            max_attempts
                        );
                        Duration::from_secs(2u64.pow(attempt - 1))
                    }
                };
                tokio::time::sleep(backoff).await;
            }

            let query_future =
                client.query_workspace(&self.workspace.workspace_id, &self.query, None);
            match tokio::time::timeout(timeout, query_future).await {
                Ok(Ok(response)) => return Ok(response),
                Ok(Err(e)) => {
                    last_error = Some(e);
                }
                Err(_) => {
                    last_error = Some(KqlPanopticonError::QueryExecutionFailed(format!(
                        "Query timed out after {} seconds on workspace '{}'",
                        timeout.as_secs(),
                        self.workspace.name
                    )));
                }
            }
        }

        // All attempts failed
        Err(last_error.unwrap_or_else(|| {
            KqlPanopticonError::QueryExecutionFailed(format!(
                "Query failed on workspace '{}' after {} attempts",
                self.workspace.name, max_attempts
            ))
        }))
    }

    /// Format a JSON value for CSV output
    fn format_csv_value(&self, value: &serde_json::Value) -> String {
        match value {
            serde_json::Value::Null => String::new(),
            serde_json::Value::Bool(b) => b.to_string(),
            serde_json::Value::Number(n) => n.to_string(),
            serde_json::Value::String(s) => {
                // Escape quotes and wrap in quotes if needed
                if s.contains(',') || s.contains('"') || s.contains('\n') {
                    format!("\"{}\"", s.replace('"', "\"\""))
                } else {
                    s.clone()
                }
            }
            serde_json::Value::Array(_) | serde_json::Value::Object(_) => {
                // Serialize complex types as JSON strings
                let json_str = value.to_string();
                format!("\"{}\"", json_str.replace('"', "\"\""))
            }
        }
    }
}