You learn about problems from employees and customers, not from an engineer
Active monitoring: Telegram/Slack alerts with clear text and priority, on-call rotation for critical events.
Monitoring and observability
Monitoring, alerting and logging
Zabbix, Grafana, Prometheus, Loki, Wazuh. Metrics, alerts and centralized logs — so you know about problems before the customer does.
When clients come to us
Three common scenarios and how we solve them.
Incidents repeat, but nobody finds the root cause
Centralized logging (Loki/ELK/Graylog) + metric correlation in Grafana — you can see what happened an hour before the incident.
Leadership asks "is everything OK?" — there's nothing to prove it
CEO/CFO dashboards: monthly uptime, incidents, SLA metrics, work delivered.
What is included
Clear boundaries: what is covered by the retainer and what is a separate project.
Included in the service
- Zabbix / Grafana / Prometheus: hardware, OS, application and service metrics
- Network monitoring: ping, SNMP, latency, jitter, link status, ISP incidents
- 1C/UNA monitoring: active sessions, locks, timings, job queues
- Logging: Graylog, Loki, ELK — centralized collection and search
- Alerts in Telegram / Slack / Email with clear text and priority
- On-call rotation and escalation — critical alerts never get "lost" at night
- Dashboards for CEO/CFO: uptime, incidents, SLA metrics by period
Not included / separate project
- Custom BI and enterprise analytics outside IT metrics
- Integrations with exotic systems without a public API — possible, but as a project
- 24/7 SOC with threat intelligence — we connect MSSP partners
What is actually included in the contract
When we sign — both sides know exactly what you pay for and what you get. These are the four planes the SLA is built on.
Coverage hours
- Business hours: Mon–Fri 08:00–20:00, L1 with guaranteed response
- On-call evenings and weekends on request
- 24/7 for critical services: mail, 1C, UNA, POS, website, VPN
- Scheduled night windows for planned work without downtime
Contact channels
- Single ticket portal with history and SLA timers
- Hotline phone for P1 incidents
- Email and Telegram bot for routine tasks
- User chat with our on-duty engineer during business hours
Priorities and response
- P1 (critical, service is down): response in 15 minutes, 24/7
- P2 (severe impact): response within 1 hour in business hours
- P3 (single-user issue): response within 4 hours
- P4 (planned work, improvement): scheduled for the coming week
Contract scope of work
- Planned maintenance: updates, backups, monitoring, patch management
- Incident management: diagnostics, recovery, postmortems
- Small projects within the retainer: up to N hours per month
- Large projects (migrations, new equipment) — separately scoped
Technologies and tools
What we work with in this area.
Zabbix
Grafana
Prometheus
Wazuh
Graylog
Loki
ELK
Alertmanager
Netdata
How we onboard a new client
1
Kickoff and brief
30–60 minutes online or in the office: we discuss the business, the key systems, and the incidents that hurt last year. We sign an NDA.
2
Technical audit and risk map
1–2 weeks. Equipment inventory, checks on servers, network, backups, security, 1C/UNA, licensing. The outcome is a report with a risk map and priorities.
3
Work plan and contract
We build the plan: what gets fixed right away (P1 risks), what in the first month, what in the first quarter. We lock the SLA, hours, channels, price. We sign the contract.
4
Onboarding and infrastructure handover
We take over the access via a secure password manager, deploy our monitoring, plug in the ticket system, train the employees. The previous admin (if any) hands off by the playbook.
5
Ongoing operations and reporting
Monthly report: incidents, planned work, new risks, recommendations. Quarterly review with the CEO/CFO: strategy, budgets, major changes.
Early detection of 1C degradation — fixed before users noticed
Distribution and logistics, Moldova
Задача
1C and UNA slowed down periodically, complaints every Monday. Root cause was found manually, post-mortem.
Решение
Monitoring of sessions, locks, operation durations and job queues. Alerts fire 15-30 min before real degradation.
- −80%"1C is slow" complaints
- 2 min (was 1-2 h)Average time to detection
- <5% after tuningFalse alerts
Frequently asked questions about this service
Can we deploy monitoring on our own infrastructure and keep it there?
Yes. We roll out Zabbix/Grafana/Loki on your side (self-hosted), hand over credentials and documentation. Or use our cluster — whichever works.
How do you monitor 1C and UNA?
Zabbix agent + dedicated scripts: active sessions, locks, operation durations, DB size, job queues, tech-process state, tempDB free space.
Do Telegram alerts get noisy?
We group by priority, apply quotas and quiet windows for non-P1. Typical load — 1-2 messages per day on mid-sized infra.
Do you keep logs long-term for compliance?
Yes. We configure retention (usually 90 days hot, up to 2 years cold in S3). Suitable for GDPR and internal audit.
Let us talk about your IT
The first meeting — 30 minutes, free, no strings attached. We will tell you whether we are the right fit and what is worth doing first.