$ cat "Enterprise 1C:Enterprise Infrastructure & Clusters.md"

Enterprise 1C:Enterprise Infrastructure & Clusters

We design, configure, and maintain server infrastructure for high-load 1C systems (ERP, Trade Management, Payroll, Corporate Finance). We eliminate document posting freezes, database locks, slow reports, and storage IOPS bottlenecks.

What We Do

  • 1C Application Server Clusters:

    • Zero Single Point of Failure (No SPOF): redundant central nodes, licensing daemons, and clustered session state persistence.
    • Workload isolation: offloading heavy background recalculations and batch reports to dedicated worker processes (rphost), shielding interactive users from slowdowns.
    • Cluster manager policies: process memory quotas, scheduled worker restarts to prevent memory leaks/fragmentation, and Session Fault Tolerance.

  • Database Engine Tuning & Administration:

    • PostgreSQL / Postgres Pro: Linux kernel parameter optimization (sysctl), shared_buffers, dirty page write tuning; autovacuum calibration for 1C transaction patterns; online_analyze and plantuner extensions.
    • Microsoft SQL Server: AlwaysOn Availability Groups, memory caps (max server memory), parallelism settings (MAXDOP, cost threshold), multi-file tempdb layouts, and trace flags.
    • Storage I/O Segregation: dedicated NVMe pools for transaction logs (WAL / LDF), temporary tables (tempdb), and database data files.

  • Diagnostics & Technological Log (TJ):

    • Continuous TJ collection and parsing (logcfg.xml): isolation of lock contention timeouts (TTIMEOUT), deadlocks (TDEADLOCK), exceptions (EXCP), and slow queries (CALL).
    • Real-time observability (Zabbix, Prometheus, Grafana): tracking disk queue depths (await), RAM usage, license seats, and transaction response times.
    • Bottleneck isolation: identifying unoptimized code and providing actionable recommendations to 1C developers (missing indexes, managed locks).

  • Backups & Disaster Recovery (DRP):

    • Continuous WAL archiving via pg_probackup for near-zero RPO.
    • Automated test restorations to an isolated sandbox to regularly verify backup integrity.
    • Formalized Disaster Recovery Plans (DRP) targeting RTO under 30 minutes.

Standard Enterprise Cluster Architecture

               [NLB / HAProxy TCP Load Balancers]
                             |
         +-------------------+-------------------+
         |                                       |
+--------v----------------+             +--------v----------------+
|  1C App Server 01 (Master)            |  1C App Server 02 (Standby)
|  - Primary Cluster Node |             |  - Backup Central Node  |
|  - Interactive Sessions |             |  - Background & Reports |
+--------+----------------+             +--------+----------------+
         |                                       |
         +-------------------+-------------------+
                             |
         +-------------------+-------------------+
         |                                       |
+--------v----------------+  Sync WAL   +--------v----------------+
|  Postgres Pro Master    |============>|  Postgres Pro Standby   |
|  - NVMe Data / NVMe WAL |  Streaming  |  - Patroni HA Agent     |
|  - Patroni HA Agent     |             |  - Read-Only Analytics  |
+-------------------------+             +-------------------------+

Technology Stack

  • Platforms: 1C:Enterprise 8.3 (ERP, Trade Management, Payroll, Corporate Finance).
  • Databases: PostgreSQL, Postgres Pro (Standard / Enterprise), Microsoft SQL Server.
  • Clustering & HA: Patroni, pg_probackup, AlwaysOn Availability Groups.
  • Load Balancing: HAProxy, Keepalived, NLB.
  • Operating Systems: Linux (Ubuntu, Debian, Astra Linux, RED OS), Windows Server.
  • Monitoring: Zabbix, Prometheus, Grafana, 1C Technological Log.

Results

  1. Responsive Performance: transactions and documents post without lock timeouts or queues, even during month-end closes and peak hours.
  2. High Availability: application server or database node failures do not interrupt business operations.
  3. Data Integrity: continuous transaction log archiving and validated fast recovery procedures.
  4. Transparent Visibility: real-time tracking of database latency, storage queues, and vital 1C performance metrics.
Discuss Project