ThemisDB v1.8.1-rc1

Release Candidate v1.8.1-rc1 (minor bugfix RC).

Highlights

• Former beta modules graduated to production-ready: api, content, geo, gpu, process, sharding
• Docker runtime paths standardised to Linux FHS (/etc/themis/config, /var/lib/themis/data, /var/log/themis)
• Spurious startup warnings downgraded to INFO (PIIDetector, gRPC stubs)
• CONFIG sig/data directory layout with backward-compat symlinks
• themisctl v0.0.2 — unified management CLI (health, version, query, get, put, delete, config, branch, snapshot, admin, repl)
• VERSION/RELEASE_TYPE set to 1.8.1-rc1 / rc

Validation

• Local production-readiness targeted gates completed
• Sharding focused test + benchmark evidence: artifacts/production-readiness/20260404_*

Docker

docker pull themisdb/themisdb:v1.8.1-rc1
docker pull themisdb/themisdb:latest

Downloads

Asset	Platform	Contents
themisdb-linux-x64-v1.8.1-rc1.zip	Linux x64	themis_server, runtime libs, config.yaml
themisdb-windows-x64-v1.8.1-rc1.zip	Windows x64	themis_server.exe, DLLs, themisctl.exe, themis-export.exe, config.yaml
themisctl-windows-x64-v1.8.1-rc1.zip	Windows x64	themisctl.exe + shell completion (bash/zsh/fish)
*.sha256.txt	—	SHA-256 checksums

Verify checksum (Linux/macOS)

sha256sum -c themisdb-linux-x64-v1.8.1-rc1.sha256.txt

Verify checksum (Windows PowerShell)

(Get-FileHash themisdb-windows-x64-v1.8.1-rc1.zip -Algorithm SHA256).Hash.ToLower()
# compare to content of themisdb-windows-x64-v1.8.1-rc1.sha256.txt

Quick start (Linux)

unzip themisdb-linux-x64-v1.8.1-rc1.zip
chmod +x ./bin/themis_server
LD_LIBRARY_PATH=./bin ./bin/themis_server --config=./config/config.yaml --data-dir=./data

Quick start (Windows)

Expand-Archive themisdb-windows-x64-v1.8.1-rc1.zip
cd themisdb-windows-x64
.\bin\themis_server.exe --config=.\config\config.yaml --data-dir=.\data

Related

• PR: #4442

ThemisDB v1.8.0

What's Changed

• docs: Complete SYSTEMATISCHER_REVIEWPLAN.md verification with Option 3 approach (98 items verified, all phases complete) + final report + issue templates by @Copilot in #408
• feat: Retroactive release building system with Git Flow integration and commit support by @Copilot in #413
• docs: Fix CMAKE_MODULAR_ARCHITECTURE.md documentation-source code gap by @Copilot in #414
• Refactor http_server.cpp: Complete foundation with structure, reference implementation, integration guides, and issue templates by @Copilot in #411
• Revert "[WIP] Implement LLM model loading from RocksDB Blob Store" by @makr-code in #404
• [DOCS] Create comprehensive src/ directory structure documentation (100% coverage) by @Copilot in #416
• Add comprehensive examples documentation with index and quickstart guides by @Copilot in #418
• Create Tools Documentation Hub and Usage Guides by @Copilot in #415
• [DOCS] CRITICAL: Correct plugin documentation - acceleration IS implemented by @Copilot in #417
• docs: Documentation cleanup investigation and Phase 1 link verification complete by @Copilot in #444
• Refactor: Extract ChangefeedApiHandler from http_server.cpp by @Copilot in #447
• [REFACTOR] Extract Cache Operations into CacheApiHandler by @Copilot in #446
• Refactor: Extract ContentApiHandler from HttpServer (~900 lines) by @Copilot in #448
• docs: Complete documentation archival system implementation with CHANGELOG and completion summary by @Copilot in #443
• Refactor: Extract EntityApiHandler from http_server.cpp (~880 lines) by @Copilot in #449
• refactor: Extract IndexApiHandler from http_server.cpp by @Copilot in #451
• [REFACTOR] Extract GraphApiHandler from http_server.cpp by @Copilot in #450
• refactor: Extract PromptApiHandler from HttpServer by @Copilot in #454
• Refactor PolicyApiHandler - Extract Ranger policy operations from http_server.cpp by @Copilot in #453
• Extract query operations from http_server.cpp into QueryApiHandler by @Copilot in #455
• REFACTOR: Extract monitoring operations into MonitoringApiHandler by @Copilot in #452
• REFACTOR: Extract transaction operations into TransactionApiHandler by @Copilot in #458
• REFACTOR: Extract spatial operations into SpatialApiHandler by @Copilot in #456
• REFACTOR: Extract time series operations into TimeSeriesApiHandler by @Copilot in #457
• Refactor vector operations into VectorApiHandler by @Copilot in #459
• Refactor: Extract WAL replication operations into WALApiHandler by @Copilot in #460
• Refactor PITR implementation to use RocksDBWrapper abstraction by @Copilot in #461
• Fix HTTP server type mismatch, integrate DiffApiHandler, and document incomplete cpp-httplib migration by @Copilot in #463
• Add independent Health/Error service on alternate port (9090) by @Copilot in #464
• Restructure README and CHANGELOG for GitHub best practices by @Copilot in #465
• Fix 11 compilation errors: missing macros, interface mismatches, missing headers by @Copilot in #469
• Research: Git/GitHub/GitOps vs ThemisDB version control comparison with existing YAML analysis by @Copilot in #470
• Add Traveling Salesman Problem example demonstrating graph-based optimization by @Copilot in #471
• Refactor repository structure: clean root directory and add GitHub best-practice files by @Copilot in #472
• Complete Stages 1-2 and begin Stages 3-41: Chapters 41-40 complete, Chapter 39 Checkpoint 2 (100%), Chapters 38-1 analyzed (39 total) with 19 issue templates created - 95% analysis coverage, 100% template coverage achieved by @Copilot in #468
• Stage 3 Checkpoint 3: Expand Chapter 39 Performance Tuning with Storage, Cache, OS, and Transaction sections by @Copilot in #492
• Chapter 37 Stage 5 CP2: Expand Sections 37.1-37.3 with Technical Depth and Benchmarks by @Copilot in #495
• Expand Chapter 38 Observability & SRE - Complete All Sections with Scientific Depth by @Copilot in #493
• [Ch.36 CP2] Expand Security: TLS 1.3, MFA, RBAC/ABAC, Secrets Management by @Copilot in #496
• Stages 3-4: Complete Chapter 39 validation and Chapter 38 comprehensive expansion by @Copilot in #494
• Stage 4 Complete: Expand Chapter 35 Data Modeling Patterns (All Sections) + Code Review Fixes by @Copilot in #497
• [Ch.34 CP2] Expand Query Planning, Execution & Index Selection with Scientific Depth by @Copilot in #498
• [Ch.33 CP2] Expand Normalization, Denormalization, Schema Evolution & Versioning by @Copilot in #499
• Close GPU/VRAM documentation-implementation gap: Enable Vulkan+LLM by default with Docker support by @Copilot in #500
• Add Chapter 32a: API Design & REST Principles with comprehensive REST fundamentals, HTTP methods, and resource design by @Copilot in #501
• Chapter 31: Add comprehensive REST, gRPC, GraphQL, and protocol sections by @Copilot in #502
• Expand Chapter 30: Deployment & Operations with Production-Ready Examples by @Copilot in #503
• Expand Chapter 27: Troubleshooting & Problem Resolution (+5,653 words, 8 scripts, 3 benchmarks) by @Copilot in #506
• Expand Chapter 28 AQL Reference with syntax fundamentals, operators, subqueries, transactions, and optimization by @Copilot in #505
• Expand Chapter 29: Analytics & Process Mining - Add OLAP, Process Mining Algorithms, Predictive Analytics by @Copilot in #504
• [CH26-CP2] Expand Migration & Legacy Integration sections 26.1-26.5 with 12 quality dimensions - Reviewed by @Copilot in #508
• [CH25-CP2] Expand DevOps & Infrastructure as Code chapter (1.6K→14.8K words) by @Copilot in #509
• docs: Add and verify comprehensive mermaid architecture diagrams against docs/de by @Copilot in #510
• Add thematic namespace and class overview documentation with Mermaid diagrams by @Copilot in #507
• [CH23-CP2] Expand Testing & Quality Assurance chapter with comprehensive coverage by @Copilot in #512
• Expand Chapter 22: Client Libraries & Drivers (Sections 22.1-22.6) by @Copilot in #513
• [CH24-CP2] Expand AI Ethics & Governance with GDPR/CCPA compliance, fairness metrics, and data governance by @Copilot in #511
• Fix clear-text logging of sensitive information in VCC Veritas adapter by @Copilot in #515
• chore(deps): Bump tj-actions/changed-files from 40 to 46 in /.github/workflows by @dependabot[bot] in #514
• Add explicit permissions to AFL++ fuzzing workflow by @Copilot in #516
• Add WordPress plugin for ThemisDB Compendium PDF downloads with interactive preview by @Copilot in #517
• WordPress Plugin Compatibility: Themis Brand Integration and Best Practices Documentation by @Copilot in #519
• LLM/LoRA System Analysis: Critical Gaps, Production Roadmap, Implementation Guide, GPU Acceleration, Base Infrastructure, Tests & Issue Templates by @Copilot in #518
• Implement llama.cpp resource management with GPU backend integration by @Copilot in #522
• Implement RSA-SHA256 signature verification with OpenSSL for LoRA adapters by @Copilot in #527
• [LoRA] Implement CPU-based LoRA training system with verified gradients and future phase templates (including QLoRA) by @Copilot in #528
• Implement token sampling strategies using llama.cpp API by @Copilot in #526
• Investigation: LLM/LoRa production-readiness gaps and ThemisDB integration analysis by @Copilot in https://github.com/ma...

ThemisDB v1.3.4

ThemisDB v1.3.4

Source Code Archive

Download the source code archive and verify its integrity:

SHA256 Checksum:

$(cat themisdb-1.3.4-source.zip.sha256)

Building from Source

# Extract archive
unzip themisdb-1.3.4-source.zip
cd themisdb-1.3.4

# Build with CMake
cmake -B build -S . -DCMAKE_BUILD_TYPE=Release
cmake --build build

Release Notes

See docs/de/archive/RELEASE_NOTES_v1.3.4.md for detailed release notes.

---

Commit: f840b2d60ca930cdc14e609f7bca14ac2805ba8f

ThemisDB v1.3.0 - Keep Your Own Llamas

ThemisDB v1.3.0 - Native LLM Integration

Release Date: 20. Dezember 2025
Code Name: "Keep Your Own Llamas"

---

🎉 Overview

ThemisDB v1.3.0 brings native LLM integration with embedded llama.cpp, enabling you to run AI/LLM workloads directly in your database without external API dependencies. This release introduces a complete plugin architecture, GPU acceleration, and enterprise-grade caching for production LLM deployments.

"ThemisDB keeps its own llamas." – Run LLaMA, Mistral, Phi-3 models (1B-70B params) directly in your database.

---

🚀 Major Features

🧠 Embedded LLM Engine (llama.cpp)

• Native Integration: llama.cpp embedded as local clone (not committed to repo)
• Model Support: GGUF format models (LLaMA 3, Mistral, Phi-3, etc.)
• Inference Engine: Full tokenization, evaluation, sampling, and detokenization pipeline
• Memory Management: Lazy model loading with configurable VRAM budgets

⚡ GPU Acceleration

• CUDA Support: NVIDIA GPU acceleration with 100x speedup vs CPU
• Metal Support: Apple Silicon optimization
• Vulkan Support: Cross-platform GPU backend
• Automatic Fallback: Graceful degradation to CPU when GPU unavailable

🧩 Plugin Architecture

• LlamaCppPlugin: Reference implementation for llama.cpp backend
• ILLMPlugin Interface: Extensible plugin system for custom LLM backends
• Plugin Manager: Centralized management with lifecycle control
• Hot-Swappable: Load/unload models and LoRA adapters dynamically

🗃️ Advanced Model Management

• Lazy Loading: Ollama-style on-demand model loading (2-3s first load, instant cache hits)
• Multi-LoRA Manager: vLLM-style support for up to 16 concurrent LoRA adapters
• Model Pinning: Prevent eviction of critical models from memory
• TTL Management: Automatic model eviction after configurable idle time (default: 30 min)

💾 Enterprise Caching

• Response Cache: Semantic caching for identical queries (70-90% cost reduction)
• Prefix Cache: Reuse common prompt prefixes across requests
• Model Metadata Cache: TBB lock-free cache for 10x faster metadata access
• KV Cache Buffer: Shared read-only buffers for 70% memory savings

🔧 Build & Deployment

• Windows/MSVC Support: PowerShell build script with Visual Studio 2022
• MSVC Fixes: /Zc:char8_t- compiler flag for llama.cpp compatibility
• Docker BuildKit: Flexible llama.cpp source (local or git clone)
• Offline-First: Root vcpkg standardization for reproducible builds

📚 Documentation

• Consolidated Guides: Root README + 17 specialized LLM docs (380 KB)
• Integration Paths: Local clone approach (no git submodules)
• API Specifications: HTTP REST + gRPC binary protocols
• Client SDKs: Python, JavaScript, Go, Rust, Java, C# examples

---

📦 What's Included

Binary Artifacts

• themis_server.exe (10.2 MB) - Windows x64 Release
• llama.dll (2.2 MB) - llama.cpp inference engine
• ggml*.dll (1.4 MB) - GGML computation kernels

Source Components

• 23 LLM Headers (include/llm/)
• 23 LLM Implementations (src/llm/)
• 17 Documentation Guides (docs/llm/)
• 8+ Test Suites (tests/test_llm_*.cpp)
• PowerShell Build Script (scripts/build-themis-server-llm.ps1)

---

🔄 Breaking Changes

⚠️ llama.cpp Submodule Removed

• Before: external/llama.cpp as git submodule
• After: Local clone in project root (excluded via .gitignore/.dockerignore)
• Migration: Clone llama.cpp locally: git clone https://github.com/ggerganov/llama.cpp.git

⚠️ vcpkg Standardization

• Before: Multiple vcpkg locations (external/vcpkg, ./vcpkg)
• After: Single root ./vcpkg with VCPKG_ROOT standardization
• Migration: Set VCPKG_ROOT=C:\VCC\themis\vcpkg (or your path)

⚠️ Docker Build Context

• Before: external/ copied into build context
• After: external/ excluded; use BuildKit --build-context for llama.cpp
• Migration: See Docker build commands below

---

🛠️ Installation & Upgrade

Windows (MSVC)

# Clone repository
git clone https://github.com/makr-code/ThemisDB.git
cd ThemisDB

# Clone llama.cpp locally (required for LLM support)
git clone https://github.com/ggerganov/llama.cpp.git llama.cpp

# Build with LLM support
powershell -File scripts/build-themis-server-llm.ps1

# Verify build
./build-msvc/Release/themis_server.exe --help

Docker (with LLM)

# With local llama.cpp clone
docker buildx build \
  --build-arg ENABLE_LLM=ON \
  --build-context llama=./llama.cpp \
  -t themisdb:v1.3.0-llm .

# Without local clone (git clone in Docker)
docker buildx build \
  --build-arg ENABLE_LLM=ON \
  -t themisdb:v1.3.0-llm .

Linux/WSL

# Clone and build
git clone https://github.com/makr-code/ThemisDB.git
cd ThemisDB
git clone https://github.com/ggerganov/llama.cpp.git llama.cpp

cmake -B build -DTHEMIS_ENABLE_LLM=ON
cmake --build build -j$(nproc)

./build/themis_server --help

---

📖 Quick Start

1. Download a Model

# Example: Mistral 7B Instruct Q4 (~4GB)
mkdir -p models
cd models
wget https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.2-GGUF/resolve/main/mistral-7b-instruct-v0.2.Q4_K_M.gguf

2. Configure LLM

# config/llm_config.yaml
llm:
  enabled: true
  plugin: llamacpp
  model:
    path: ./models/mistral-7b-instruct-v0.2.Q4_K_M.gguf
    n_gpu_layers: 32  # GPU offload layers
    n_ctx: 4096       # Context window
  cache:
    max_models: 3
    max_vram_mb: 24576  # 24 GB

3. Start Server

./themis_server --config config/llm_config.yaml

4. Run Inference (HTTP API)

curl -X POST http://localhost:8765/api/llm/generate \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "What is ThemisDB?",
    "max_tokens": 512,
    "temperature": 0.7
  }'

---

🎯 Performance Benchmarks

GPU vs CPU (Mistral-7B Q4, RTX 4090)

Operation	CPU (20 cores)	GPU (CUDA)	Speedup
Model Load	2.8s	2.1s	1.3x
Inference (512 tokens)	32s	0.3s	107x
Throughput	16 tok/s	1,700 tok/s	106x

Memory Usage (with Caching)

Feature	Memory	Savings
Base Model (Mistral-7B Q4)	4.2 GB	-
+ Response Cache	4.5 GB	85% query cost
+ Prefix Cache	4.6 GB	40% latency
+ KV Cache Sharing	3.1 GB	70% memory

Lazy Loading Impact

Scenario	Cold Start	Warm Cache	Benefit
First Request	2.8s	-	-
Subsequent Requests	-	~0ms	Instant
After TTL Expiry	2.8s	-	Auto-reload

---

🔒 Security Considerations

• Model Files: Store models outside web root with proper permissions
• API Authentication: Enable Bearer Token (JWT) authentication in production
• Rate Limiting: Configure per-user quotas for inference requests
• Resource Limits: Set max_vram_mb and max_models to prevent exhaustion
• Audit Logging: All LLM operations logged for compliance

---

📊 Known Limitations

1. Windows DLL Export Limit: Use static build (THEMIS_CORE_SHARED=OFF) to avoid 65k symbol limit
2. GPU Memory: Requires sufficient VRAM for model + overhead (~100 MB CUDA)
3. Model Format: Only GGUF format supported (llama.cpp v2+)
4. Concurrent Requests: Limited by available VRAM and KV cache size
5. Docker BuildKit: Requires Docker 19.03+ and BuildKit enabled

---

🐛 Bug Fixes

• Fixed MSVC char8_t compilation errors in llama.cpp via /Zc:char8_t- flag
• Resolved vcpkg path conflicts between external/vcpkg and root ./vcpkg
• Corrected Docker .dockerignore to exclude llama.cpp/ from build context
• Removed circular submodule dependencies (docker/tmp/openssl, external/llama.cpp)
• Fixed CMake generator/architecture issues on Windows (requires -A x64)

---

📝 Deprecations

• Git Submodules for llama.cpp: Deprecated in favor of local clone approach
• external/vcpkg: Deprecated in favor of root ./vcpkg location
• Manual llama.cpp Setup: Use scripts/setup-llamacpp.sh or PowerShell build script

---

🔮 Roadmap (v1.4.0)

• Streaming Generation: Server-Sent Events (SSE) for real-time responses
• Batch Inference: Process multiple requests in single forward pass
• Distributed Sharding: Multi-node LLM deployment with etcd coordination
• vLLM Plugin: Native vLLM backend for PagedAttention and continuous batching
• Model Replication: Raft consensus for cross-shard model synchronization
• Advanced Quantization: Support for AWQ, GPTQ, and custom quantization schemes

---

🙏 Acknowledgments

• llama.cpp Team: For the incredible inference engine (MIT License)
• GGML: For efficient tensor operations on CPU/GPU
• HuggingFace: For GGUF model hosting and community
• ThemisDB Contributors: For testing, feedback, and documentation improvements

---

📚 Documentation

• LLM Integration Guide: docs/llm/LLAMA_CPP_INTEGRATION.md
• Plugin Development: docs/llm/README_PLUGINS.md
• Architecture Review: docs/llm/INTEGRATION_REVIEW_AND_SEQUENCE.md
• HTTP API Spec: docs/llm/HTTP_API_SPECIFICATION.md
• Docker Deployment: DOCKER_DEPLOYMENT.md
• Build Guide: docs/build/README.md

---

📞 Support & Community

• GitHub Issues: [Report bugs or request features](https://git...

ThemisDB v1.0.0

ThemisDB v1.0.0

Source Code Archive

Download the source code archive and verify its integrity:

SHA256 Checksum:

$(cat themisdb-1.0.0-source.zip.sha256)

Building from Source

# Extract archive
unzip themisdb-1.0.0-source.zip
cd themisdb-1.0.0

# Build with CMake
cmake -B build -S . -DCMAKE_BUILD_TYPE=Release
cmake --build build

Release Notes

See docs/de/archive/RELEASE_NOTES_v1.0.0.md for detailed release notes.

---

Commit: 2103e46d6de0b899ad0aa2c8d276551d66b27e9d