How we implemented GLPI
The action takes place in one of the countries of Central Asia, I will not name the country, the company and the area of its activity. Hopefully intrigued. Although, in fact, it is irrelevant. I would like to tell you how we started from scratch and built a fairly adequate and manageable structure in which everything is laid out on the shelves. So this story is not so much about the capabilities of the platform, but about the experience of using it as a departmental management system in real conditions.
The company in which our gallant IT department works is a subsidiary of the Corporation belonging to the Fortune 500. In our own country of the Corporation, there is absolutely everything for us to see: us IT strategy, IT policy-standards-procedures, our own IT company with a full profile, your Tier 3 data center , direct contacts and pricing agreements with major manufacturers of IT equipment and software, fast and cheap equipment and service providers, clever integrators, intelligent implementation consultants, etc. etc. - in general, there is paradise. They are expecting and demanding the same level from us, but again, we work in Central Asia and we have a solid construction here.
At first, the Company in terms of IT managed with the services of subcontractors. As a freelance IT specialist, I served offices of 2-3 subcontractors. It ended with one of my clients who bought all my time and seconded me to a remote site, where they built an industrial facility for the Company. After working there for 3 years, I learned that the Company decided to take a full-time IT coordinator in the country. By that time I already had 4 coordinators from different offices and structures of the Center, so I wouldn’t transfer one more. He went to work at the Company. The infrastructure grew very quickly, the Company's staff grew even faster, and all the work on IT was now performed by contractors. In addition to IT contractors, I had subordinates, over time we were assigned to the department, which I now lead.
Why am I telling all this and where is GLPI ? I just love to tell stories with details. You can safely skip a couple of paragraphs.
')
Our first technical support contractor was a branch of a British IT company. I will call them British. They had helpdesk that was rolled back on many clients all over the world, written software for them, trained operators, disciplined engineers, and strict rules regarding the design of applications.
The second contractor was a local company that unexpectedly won the tender with dumping prices. Let them be temporary workers - no offense, they knew that they signed up for a short intermediate period in the hope of expressing themselves and without any special chances to win the next long-term contract. The technical conditions of the contract required the presence of the application registration system, respectively, the company bought some inexpensive software, planted the girl with knowledge of the English language, recruited technicians and engineers - and, despite the apocalyptic predictions of the British, we survived the first 3 months of the contract, and then 6 months of renewal.
Everyone tried very hard. There was a lot of work: while the temporary worker repulsed the routine, we, that is, the company's own IT staff, few, coordinated the arrangement of the new office (4 floors, about 2 thousand square meters, about 400 jobs) and with minimal downtime for the main business moved and continued to work in a new place. In fact, there was no time for helpdesk, so I don’t even remember what they had there and how it worked. I also don’t recall any special problems with technical support. And we did not forget about global problems ...
And then came the third - the current - the contractor. By this time, we were already fed up with explaining to the new team what and how it worked and how it all should work, but what had to be done - we only had the final report of the previous contractor, scattered documentation on the system and we could only describe the workflows in words. The understanding came that it would be necessary to have this information at hand, so as not to repeat to people from time to time ... After all, even just a helpdesk database is a storehouse of information, and information about us. And know-how, and this is our know-how ... Why then do we not keep this information in our possession?
The contract was coming long - for 5 years - so that the new guys were determined to do everything correctly and thoroughly. And, of course, better than predecessors. Although they were also, as they say, first-time. But they said helpdesk especially at the tender stage and even wanted to develop it as a commercial service. Thanks to them, by the way, the helpdesk at an industrial facility we began to work 24x7.
Among other things, a discussion ensued about choosing a helpdesk platform. The British took their helpdesk. Helpdesk temporary worker was not even considered. Naturally, I asked the head office for information about their helpdesk. It turned out that they use CA Services Desk Manager from CA Technologies, their license for us is not translated in any way, no one is going to introduce helpdesk with us in the foreseeable future. And in general, they have a large-scale reorganization. But at least we now know what we want! So we make a call to the exclusive seller of software in the country. Here it begins, as usual: what exactly do you need, where is the technical project, why exactly that ... Well, read the product descriptions on the CA website, made up the wish list. After some time, we get a commercial offer - there were so many zeros that the sellers re-checked their number with their supplier several times. And this is only for the license, they said, and it is not known what the implementation will cost and who will implement it. Usually, they said, the license is 1/4 of the total cost of the project - but it’s generally difficult to promise something. Well, therefore, not destiny ... And the time is gone. And the operators write tickets in Excel.
I climb in Google. I learn about GLPI . I put on a free car - it works, well, never mind yourself! I show new contractors ... That's why I don’t like engineers, it’s because they see problems in everything and are not able to share the joy of their colleagues. Of course, my GLPI is some kind of left, but they found the system - this is the system. True, paid, but were we anyway ready to buy CA? But she can do this, that, the fifth, and the tenth. And I already feel sorry, this is what I, in vain put GLPI? In addition, it was not so much a pity for the money, but I did not want to get involved in a tedious system of approvals and procurement through a tender. And GLPI is already up and running. I read it, picked it up in the settings - it turns out, GLPI can do it all + screw up the inventory , which I also really needed. The engineers went to read about their software and, of course, it was found that you can easily purchase an inventory module. In general, I won. While they were looking for additional counter-arguments, I taught the operators to use GLPI and, over time, they stopped even (at least with me) writing applications to Excel in parallel. But all the same, at first all the applications were neatly written down on paper.
In general, for about 2 years we have worked a bunch of GLPI + OCS. All this time, only helpdesk operators used the system - they registered applications 24x7, occasionally the chief contractor engineer, because he also considered this approach to be correct, and so did I. During this time, I managed to transfer the OCS + GLPI servers to a virtual machine with the transition to a new version of both Ubuntu and OCS and GLPI. OCS gradually arranged agents through GPO, and I used an administrative resource, forcing technicians to enter accounting inventory numbers into OCS and GLPI. In addition, I insisted that the contents of the warehouses be entered into the GLPI database. Thus, we got a bunch of user / location-serial number-inventory inventory.
Using the Data Injection plugin, I added radios, radios, and the rest of the accountable reporting to GLPI. Right in the radio station added scans permissions, licenses, forms and letters. I basically did not make changes to the data structures and simply stuck everything in the category Devices (Devices (?)). I issued the licenses and permits with validity terms as contracts - and received ready-made planned notifications about the expiration of the validity of such documents. Convenient, but no one except me needed it. They can always ask and they will answer. You just have to remember who to ask about.
The system helped out a great deal during a checkout arranged by head-IT. It was almost the first time that someone read our tickets with interest, requested samples by criteria, looked at the statistics for closing requests, selectively checked how the user-tag-serial link works. For this part, there were no complaints.
After checking, I relaxed, so for a while nothing substantial happened in this area.
The most interesting thing started when the Company suddenly decided to hire almost all of the contractor’s personnel directly - it was necessary to urgently improve the statistics on the nationalization of the Company under an agreement with the government. The IT department suddenly grew many times, the usual way in which the Company's own IT equipment was exercising the overall management of the contractor’s staff broke down, and now everyone is in an equal position. Inevitably, floatings appeared about the uneven distribution of the load between veterans and the new staff, the Company's staff and senior contractor engineers - those who did not pass to us. New centers of power were formed, opposition was formed. Previously, just that - you could go and make a complaint to the contractor and their management quickly resolved issues, including the replacement of personnel. And now I am responsible for them.
Introduced rotation of staff within the team. Two tasks were solved: 1) interchangeability within the team, 2) lack of specialists at remote sites. A common calendar or duty schedule would help a lot here, but I didn’t find anything like that in GLPI or for GLPI.
Inside the team painted responsibilities. The list of duties we have developed gradually. I have already mentioned the rotation, so, we tried to ensure that each role (or duty) was performed by at least two people in different shifts, so that they would change each other. Responsibilities can be reflected in the system either through groups of technicians or through entities (we use the English interface, I do not know how this word is translated in the Russian version) - this is something like OU, an organizational unit in the active directory.
Our work consists of the inevitable routine and interesting projects. Routine should be as little as possible, and the projects - the more the better. Part of the projects I suggested based on the needs of the office, part of the people put forward himself. We agreed that projects and necessary resources (time, people, materials, money, ...) will be coordinated with me in advance. If you don’t want to do a routine, take the projects I need. No projects - do the routine. GLPI planning was used to divide the time between doing the routine and doing the projects or scheduled maintenance. Now the calendar shows who is currently busy. To control the execution of projects in GLPI, there was a simple but nice project management plugin with drawing Gantt charts. The project can hang a lot of system objects. We are just starting to enter projects into GLPI - and, again, this is just what I need, because there are many people, many projects, and sclerosis is progressing.
In order to achieve a fair distribution of applications, the team was assembled in one place: before this, the equipment had been distributed among different office blocks. At the same time, they slightly lost in response time to applications, but achieved their own. Agreed that all applications and tasks are distributed through helpdesk - so nobody was offended. We established a simple queue when assigning applications to free technicians. Analyzed the data on the number of completed applications per unit of time - there were arguments for a serious conversation with some technicians. And, finally, it was possible to force technicians to register all applications. It was not possible yet to force them to close orders independently: it continues to do helpdesk. In turn, the helpdesk was explained that now it would not be possible to write down requests for a piece of paper and deposit them into the system at the end of the shift - otherwise the opening and closing times of the request would coincide. Helpdesk in general with us - well done. One had only to show them how to use templates for automating and standardizing the most frequent applications - and they became the most appreciative GLPI adherents and the best assistants in the further development of the system. It seems to me that with minimal system prompts, operators began to feel much more confident. I plan to use the time available in the help of the helpdesk to look after the NMS, to distant support, yes to anything: I need to remove them from the risk of staff cuts during the crisis, so that even the thought of the authorities did not occur ...
By the way, there are more applications. And they have become much better decorated. If earlier the application looked like this: something happened to a user like this - the application is open / the application is closed, now for a good half of the applications it is enough to select the category of the application and fill in the required fields from the template, the application is assigned to a specific technician or technicians group in accordance with with their specialization, in the description there is a hint for the operator. This may be a typical description of the problem, a link to a document in the knowledge base, a list of information that needs to be clarified with the user, the procedure for the actions of the operator - whatever. I do not even know whether to give out all our brilliant ideas for free. On the one hand, I really want to brag, but on the other, thoughts about monetization creep in ... However, as it turns out, the services market is already narrower than we imagined quite recently, I’ll explain why below.
Thanks to a clear demonstration on examples from applications, people understood why SLA is needed, how escalation is carried out, and all this is done automatically through categories and application templates. And now I have a live and working SLA list, that is, I have established standards for the performance of certain works.
We have not yet mastered the business rules - the meaning is clear, the functionality is needed, but so far the criteria do not pass as we want. Not scary, we will understand.
Draw a network map in the plugin Network architecture . Handsomely. For some reason, the center sometimes wants to know with accuracy to the port that it is connected to. I definitely need such a card to be, so that it lies in a certain place and is accessible from everywhere (access, of course, we control). For the Center, for presentations, for team interaction, for completeness of coverage of the system, and just in case. And so our networkers have cards in Visio, in Cisco Prime - but it is necessary to request someone, wait until they answer, but you know these engineers - they are all negative.
Budget - you can paint a budget up to the cost of individual pieces of equipment or materials or services. With the binding of financial information to equipment and contracts. As for me, we have a very difficult budget cycle. Approximately every 2 months, different instances require reports with varying degrees of detail. I will definitely write down in detail my budget for next year in GLPI. So far there are only 2 articles: CAPEX and OPEX.
Contacts - here I entered all the phone numbers / addresses, the search for which used to take time. Up to restaurants where we sometimes order food with delivery. People always smile, but in fact we save time and make our lives easier.
Contracts - everything is clear. Any information about contracts. With expiration reminders. Linked to equipment - I have already mentioned that I use this category for licenses to operate radio equipment, import permits, and so on.
Problems - here I enter everything that is vaguely disturbing and does not yet have an immediate solution. Then you can attach the problem to requests to be tied to specific events. You can bind the problem to the project. You can mention it in the budget for the next period.
It seems that nothing of the functionality used by us is forgotten.
Summing up our experience, - The control system apparently does not need a technician / engineer. They categorically do not want to share knowledge, and even more so to lay out the information in writing. Even if someone suddenly shared - why do we need a knowledge base at all, if he already knows everything, and if he doesn’t know, then this is a problem of the authorities, and if he suddenly wants to figure it out, Google will ask. Accounting applications is also not needed, in any case, it recycles, and the rest is not working out and let the authorities understand.
The people recognize only the regulations, only the instructions, only the control - and all this is in my different form placed in GLPI. It seems to me right for management to keep feedback open, stick to common sense, be predictable and be fair - but this is not necessary. The main thing to remember is that someone promised - and this, again, can be fixed in GLPI.
I did not need a management system as the head of the department while the bulk of the work was performed by outsourcing. Outsourcing can always be put in a dependent position. If you need to change the service provider, of course, problems arise, but if there are financial levers of pressure, this is not essential. Some people even like such problems. Some even deliberately create such problems.
, . «» , «» «» . , . . . «» , , : «, … - … freeware opensource? enterprise? enterprise-grade. GLPI .» ! , «», , , «think outside the box»…
, , — GLPI , , GLPI , — , , «»… , GLPI , . GPL2/GNU, — .
GLPI , . , , , . GLPI ? . GLPI . , .
, , -, . , . , . .
.
The company in which our gallant IT department works is a subsidiary of the Corporation belonging to the Fortune 500. In our own country of the Corporation, there is absolutely everything for us to see: us IT strategy, IT policy-standards-procedures, our own IT company with a full profile, your Tier 3 data center , direct contacts and pricing agreements with major manufacturers of IT equipment and software, fast and cheap equipment and service providers, clever integrators, intelligent implementation consultants, etc. etc. - in general, there is paradise. They are expecting and demanding the same level from us, but again, we work in Central Asia and we have a solid construction here.
At first, the Company in terms of IT managed with the services of subcontractors. As a freelance IT specialist, I served offices of 2-3 subcontractors. It ended with one of my clients who bought all my time and seconded me to a remote site, where they built an industrial facility for the Company. After working there for 3 years, I learned that the Company decided to take a full-time IT coordinator in the country. By that time I already had 4 coordinators from different offices and structures of the Center, so I wouldn’t transfer one more. He went to work at the Company. The infrastructure grew very quickly, the Company's staff grew even faster, and all the work on IT was now performed by contractors. In addition to IT contractors, I had subordinates, over time we were assigned to the department, which I now lead.
Why am I telling all this and where is GLPI ? I just love to tell stories with details. You can safely skip a couple of paragraphs.
')
Our first technical support contractor was a branch of a British IT company. I will call them British. They had helpdesk that was rolled back on many clients all over the world, written software for them, trained operators, disciplined engineers, and strict rules regarding the design of applications.
The second contractor was a local company that unexpectedly won the tender with dumping prices. Let them be temporary workers - no offense, they knew that they signed up for a short intermediate period in the hope of expressing themselves and without any special chances to win the next long-term contract. The technical conditions of the contract required the presence of the application registration system, respectively, the company bought some inexpensive software, planted the girl with knowledge of the English language, recruited technicians and engineers - and, despite the apocalyptic predictions of the British, we survived the first 3 months of the contract, and then 6 months of renewal.
Everyone tried very hard. There was a lot of work: while the temporary worker repulsed the routine, we, that is, the company's own IT staff, few, coordinated the arrangement of the new office (4 floors, about 2 thousand square meters, about 400 jobs) and with minimal downtime for the main business moved and continued to work in a new place. In fact, there was no time for helpdesk, so I don’t even remember what they had there and how it worked. I also don’t recall any special problems with technical support. And we did not forget about global problems ...
And then came the third - the current - the contractor. By this time, we were already fed up with explaining to the new team what and how it worked and how it all should work, but what had to be done - we only had the final report of the previous contractor, scattered documentation on the system and we could only describe the workflows in words. The understanding came that it would be necessary to have this information at hand, so as not to repeat to people from time to time ... After all, even just a helpdesk database is a storehouse of information, and information about us. And know-how, and this is our know-how ... Why then do we not keep this information in our possession?
The contract was coming long - for 5 years - so that the new guys were determined to do everything correctly and thoroughly. And, of course, better than predecessors. Although they were also, as they say, first-time. But they said helpdesk especially at the tender stage and even wanted to develop it as a commercial service. Thanks to them, by the way, the helpdesk at an industrial facility we began to work 24x7.
Among other things, a discussion ensued about choosing a helpdesk platform. The British took their helpdesk. Helpdesk temporary worker was not even considered. Naturally, I asked the head office for information about their helpdesk. It turned out that they use CA Services Desk Manager from CA Technologies, their license for us is not translated in any way, no one is going to introduce helpdesk with us in the foreseeable future. And in general, they have a large-scale reorganization. But at least we now know what we want! So we make a call to the exclusive seller of software in the country. Here it begins, as usual: what exactly do you need, where is the technical project, why exactly that ... Well, read the product descriptions on the CA website, made up the wish list. After some time, we get a commercial offer - there were so many zeros that the sellers re-checked their number with their supplier several times. And this is only for the license, they said, and it is not known what the implementation will cost and who will implement it. Usually, they said, the license is 1/4 of the total cost of the project - but it’s generally difficult to promise something. Well, therefore, not destiny ... And the time is gone. And the operators write tickets in Excel.
I climb in Google. I learn about GLPI . I put on a free car - it works, well, never mind yourself! I show new contractors ... That's why I don’t like engineers, it’s because they see problems in everything and are not able to share the joy of their colleagues. Of course, my GLPI is some kind of left, but they found the system - this is the system. True, paid, but were we anyway ready to buy CA? But she can do this, that, the fifth, and the tenth. And I already feel sorry, this is what I, in vain put GLPI? In addition, it was not so much a pity for the money, but I did not want to get involved in a tedious system of approvals and procurement through a tender. And GLPI is already up and running. I read it, picked it up in the settings - it turns out, GLPI can do it all + screw up the inventory , which I also really needed. The engineers went to read about their software and, of course, it was found that you can easily purchase an inventory module. In general, I won. While they were looking for additional counter-arguments, I taught the operators to use GLPI and, over time, they stopped even (at least with me) writing applications to Excel in parallel. But all the same, at first all the applications were neatly written down on paper.
In general, for about 2 years we have worked a bunch of GLPI + OCS. All this time, only helpdesk operators used the system - they registered applications 24x7, occasionally the chief contractor engineer, because he also considered this approach to be correct, and so did I. During this time, I managed to transfer the OCS + GLPI servers to a virtual machine with the transition to a new version of both Ubuntu and OCS and GLPI. OCS gradually arranged agents through GPO, and I used an administrative resource, forcing technicians to enter accounting inventory numbers into OCS and GLPI. In addition, I insisted that the contents of the warehouses be entered into the GLPI database. Thus, we got a bunch of user / location-serial number-inventory inventory.
Using the Data Injection plugin, I added radios, radios, and the rest of the accountable reporting to GLPI. Right in the radio station added scans permissions, licenses, forms and letters. I basically did not make changes to the data structures and simply stuck everything in the category Devices (Devices (?)). I issued the licenses and permits with validity terms as contracts - and received ready-made planned notifications about the expiration of the validity of such documents. Convenient, but no one except me needed it. They can always ask and they will answer. You just have to remember who to ask about.
The system helped out a great deal during a checkout arranged by head-IT. It was almost the first time that someone read our tickets with interest, requested samples by criteria, looked at the statistics for closing requests, selectively checked how the user-tag-serial link works. For this part, there were no complaints.
After checking, I relaxed, so for a while nothing substantial happened in this area.
The most interesting thing started when the Company suddenly decided to hire almost all of the contractor’s personnel directly - it was necessary to urgently improve the statistics on the nationalization of the Company under an agreement with the government. The IT department suddenly grew many times, the usual way in which the Company's own IT equipment was exercising the overall management of the contractor’s staff broke down, and now everyone is in an equal position. Inevitably, floatings appeared about the uneven distribution of the load between veterans and the new staff, the Company's staff and senior contractor engineers - those who did not pass to us. New centers of power were formed, opposition was formed. Previously, just that - you could go and make a complaint to the contractor and their management quickly resolved issues, including the replacement of personnel. And now I am responsible for them.
Introduced rotation of staff within the team. Two tasks were solved: 1) interchangeability within the team, 2) lack of specialists at remote sites. A common calendar or duty schedule would help a lot here, but I didn’t find anything like that in GLPI or for GLPI.
Inside the team painted responsibilities. The list of duties we have developed gradually. I have already mentioned the rotation, so, we tried to ensure that each role (or duty) was performed by at least two people in different shifts, so that they would change each other. Responsibilities can be reflected in the system either through groups of technicians or through entities (we use the English interface, I do not know how this word is translated in the Russian version) - this is something like OU, an organizational unit in the active directory.
Our work consists of the inevitable routine and interesting projects. Routine should be as little as possible, and the projects - the more the better. Part of the projects I suggested based on the needs of the office, part of the people put forward himself. We agreed that projects and necessary resources (time, people, materials, money, ...) will be coordinated with me in advance. If you don’t want to do a routine, take the projects I need. No projects - do the routine. GLPI planning was used to divide the time between doing the routine and doing the projects or scheduled maintenance. Now the calendar shows who is currently busy. To control the execution of projects in GLPI, there was a simple but nice project management plugin with drawing Gantt charts. The project can hang a lot of system objects. We are just starting to enter projects into GLPI - and, again, this is just what I need, because there are many people, many projects, and sclerosis is progressing.
In order to achieve a fair distribution of applications, the team was assembled in one place: before this, the equipment had been distributed among different office blocks. At the same time, they slightly lost in response time to applications, but achieved their own. Agreed that all applications and tasks are distributed through helpdesk - so nobody was offended. We established a simple queue when assigning applications to free technicians. Analyzed the data on the number of completed applications per unit of time - there were arguments for a serious conversation with some technicians. And, finally, it was possible to force technicians to register all applications. It was not possible yet to force them to close orders independently: it continues to do helpdesk. In turn, the helpdesk was explained that now it would not be possible to write down requests for a piece of paper and deposit them into the system at the end of the shift - otherwise the opening and closing times of the request would coincide. Helpdesk in general with us - well done. One had only to show them how to use templates for automating and standardizing the most frequent applications - and they became the most appreciative GLPI adherents and the best assistants in the further development of the system. It seems to me that with minimal system prompts, operators began to feel much more confident. I plan to use the time available in the help of the helpdesk to look after the NMS, to distant support, yes to anything: I need to remove them from the risk of staff cuts during the crisis, so that even the thought of the authorities did not occur ...
By the way, there are more applications. And they have become much better decorated. If earlier the application looked like this: something happened to a user like this - the application is open / the application is closed, now for a good half of the applications it is enough to select the category of the application and fill in the required fields from the template, the application is assigned to a specific technician or technicians group in accordance with with their specialization, in the description there is a hint for the operator. This may be a typical description of the problem, a link to a document in the knowledge base, a list of information that needs to be clarified with the user, the procedure for the actions of the operator - whatever. I do not even know whether to give out all our brilliant ideas for free. On the one hand, I really want to brag, but on the other, thoughts about monetization creep in ... However, as it turns out, the services market is already narrower than we imagined quite recently, I’ll explain why below.
Thanks to a clear demonstration on examples from applications, people understood why SLA is needed, how escalation is carried out, and all this is done automatically through categories and application templates. And now I have a live and working SLA list, that is, I have established standards for the performance of certain works.
We have not yet mastered the business rules - the meaning is clear, the functionality is needed, but so far the criteria do not pass as we want. Not scary, we will understand.
Draw a network map in the plugin Network architecture . Handsomely. For some reason, the center sometimes wants to know with accuracy to the port that it is connected to. I definitely need such a card to be, so that it lies in a certain place and is accessible from everywhere (access, of course, we control). For the Center, for presentations, for team interaction, for completeness of coverage of the system, and just in case. And so our networkers have cards in Visio, in Cisco Prime - but it is necessary to request someone, wait until they answer, but you know these engineers - they are all negative.
Budget - you can paint a budget up to the cost of individual pieces of equipment or materials or services. With the binding of financial information to equipment and contracts. As for me, we have a very difficult budget cycle. Approximately every 2 months, different instances require reports with varying degrees of detail. I will definitely write down in detail my budget for next year in GLPI. So far there are only 2 articles: CAPEX and OPEX.
Contacts - here I entered all the phone numbers / addresses, the search for which used to take time. Up to restaurants where we sometimes order food with delivery. People always smile, but in fact we save time and make our lives easier.
Contracts - everything is clear. Any information about contracts. With expiration reminders. Linked to equipment - I have already mentioned that I use this category for licenses to operate radio equipment, import permits, and so on.
Problems - here I enter everything that is vaguely disturbing and does not yet have an immediate solution. Then you can attach the problem to requests to be tied to specific events. You can bind the problem to the project. You can mention it in the budget for the next period.
It seems that nothing of the functionality used by us is forgotten.
Summing up our experience, - The control system apparently does not need a technician / engineer. They categorically do not want to share knowledge, and even more so to lay out the information in writing. Even if someone suddenly shared - why do we need a knowledge base at all, if he already knows everything, and if he doesn’t know, then this is a problem of the authorities, and if he suddenly wants to figure it out, Google will ask. Accounting applications is also not needed, in any case, it recycles, and the rest is not working out and let the authorities understand.
The people recognize only the regulations, only the instructions, only the control - and all this is in my different form placed in GLPI. It seems to me right for management to keep feedback open, stick to common sense, be predictable and be fair - but this is not necessary. The main thing to remember is that someone promised - and this, again, can be fixed in GLPI.
I did not need a management system as the head of the department while the bulk of the work was performed by outsourcing. Outsourcing can always be put in a dependent position. If you need to change the service provider, of course, problems arise, but if there are financial levers of pressure, this is not essential. Some people even like such problems. Some even deliberately create such problems.
, . «» , «» «» . , . . . «» , , : «, … - … freeware opensource? enterprise? enterprise-grade. GLPI .» ! , «», , , «think outside the box»…
, , — GLPI , , GLPI , — , , «»… , GLPI , . GPL2/GNU, — .
GLPI , . , , , . GLPI ? . GLPI . , .
, , -, . , . , . .
.
Source: https://habr.com/ru/post/272843/
All Articles