A consultation and a detailed list of recommendations for further actions are needed.
What is on the 1C side:
UTP 1.2.75, standard forms
Works under Windows Server 2012
The architecture is distributed across separate servers:
- terminal for user work
- separate server for the 1C program
- separate server for IIS
- separate server for SQL
Active users during peak periods - 30-35, which is quite low.
The configuration itself is "written-rewritten", most objects have been refined for the company's business processes.
Refinements over the last 8 years have been made by many developers, mostly freelancers.
As a result - during peak periods, significant system hangs, inability to work.
Who is on the specialists' side:
- an administrator who "oversees" the server's operation in general. However, they have no experience in optimizing 1C itself and cannot identify which bottlenecks cause such hangs during peak periods.
- a 1C developer who has been working with us for a relatively short time. If specified which fragment of code needs optimization - they will perform such optimization. However, they do not have enough skills to find these bottlenecks in "live" operation mode. Of course, they use SQL profiler and regularly catch, identify, and optimize heavy queries. However, the overall effect from this is minimal, as hangs do not always accompany the presence of heavy queries.
What is needed:
- conduct a general review of the system (access will be provided after the contractor is selected)
- provide recommendations for the admin - what exactly on the server side should be fine-tuned or what aspects of server operation should be focused on
- provide recommendations for the 1C developer: a sequence of actions on how to catch and identify bottlenecks that need optimization during peak periods
- recommend the most optimal monitoring system for our situation, if such is justified. For example, something like https://infostart.ru/1c/articles/1492368/. Accordingly, along with the recommended system, provide instructions for the developer on how to use it to catch bottlenecks in the code that need optimization. Also, provide instructions for the administrator on deploying and configuring such a monitoring system on our server.
What is not needed:
- please do not write recommendations like "When 1C hangs - you need to ask users which reports or documents they are running". Everyone understands this, it is obvious, but the team is 90% female, everyone does everything, everyone runs everything, and during working hours, stopping everyone to ask who among the 35 people ran what - will not work. Even if it does work (such attempts have been made), it gives no clues.
- it is also not worth writing about regulations - of course, during peak periods all regulations are turned off, each of them is launched in turn and its impact on the overall load of 1C is monitored. The result from this is zero. The regulations are not perfect, but if the problem were only in them, it would have been solved long ago.
- also, please do not write that our programmer is not sufficiently qualified. If you are a specialist, that is exactly why we need consultations, because the programmer is not sufficiently qualified in optimization issues for 1C. If you are a pigeon who likes to criticize others, then our cooperation is unlikely to be possible.
