Ask for help: how do I get CentOS7 to find the block file that stores the mapping?

Business scenario and historical background

there is currently a customer with two servers and a wave storage system (AS1100H mainframe + expansion cabinet, about 20T of space) running a business system.

at the beginning of the project, due to financial problems, the customer only had one server, and there was no optical communication, directly through the left and right ports of the optical fiber to card 1 and card 2 of the server HBA card. Through the wave, their storage management software (Inspur SANTricity) set up a LUN group, and then don"t know how to set it in that server, but there is a block file called "yellow" under the / dev of the server that is running now.

then my predecessor suddenly left, and I didn"t know the customer had such a business system. Now the company says that the customer has added another server, but there is still no optical communication, and the customer requires the new server to be cold standby.

I haven"t touched the storage.

what I think, I don"t know if it"s right or not.

I first opened the storage system with SANTricity to have a look. Originally, there was only one group, and the host mapped one, binding the two optical fibers corresponding to the WWN number of the HBA card. Then there are no other settings.
I think because it is all for this group and also for this block file, my plan is not to add a new host, but to tie in the WWN numbers of the two HBA of the new host.

the wiring diagram is as follows:

configuration is 

Storage
  |-> Groups(1)
         |-> Host(system123)
              |-> type: Linux(DM-MP)
              |-> Interface type: Fibre Channel
                     |-> Host port identifier: HBA1WWN
                     |-> Alias: port1
                     |-> Host port identifier: HBA2WWN
                     |-> Alias: port2
                     |-> Host port identifier: HBA1WWN
                     |-> Alias: cb-srv-port1
                     |-> Host port identifier: HBA2WWN
                     |-> Alias: cb-srv-port2

problems encountered

after starting with this setting, the new server still cannot scan the storage space. After entering the system, centos7 searched / dev and couldn"t find the block yellow (yellow is the name of this block seen under the old host / dev/).

what should I do?

I touch storage for the first time in my life. I basically don"t know anything, and I have no idea. All I know for sure is that I haven"t done anything yet. I have also heard my colleagues say that it is best not to mount this block at the same time. So all I want to do now is to find out the block of yellow.

as for how to do cold backup, I took a look at the configuration of the old server. The system they bought seems to back up the business system to the corresponding partition of yellow, so if the old server goes down, the new server can run as long as mount.

the customer is an agency unit, they have outsourced everything, this business system is done by another company, our company is the outsourced operation and maintenance, now there is no corresponding person, so push me out. Exercise.

I also hope that the elders will give us a lot of advice!

Centos7
Nov.15,2021

I think I get it.
although HBA cards of two servers can be registered on the same host, by default, they are only used by the same machine as the original server, so it is no problem to connect. However, during the initial scan, because the request is sent from another machine, the verification result is different, and there is no additional partition authorization, so the corresponding partition cannot be scanned.

the previous one mapped the entire array of space to the old server exclusive, so this storage has only one old host partition authorization. Partition authorization between hosts is isolated and independent, but a volume can only map at most one independent host or packet, so stop the current service and map the authorization to the second host. The specific steps are as follows:

  1. deploy the corresponding service on a cold standby server
  2. shut down the current service
  3. exclusive server shutdown
  4. right-click target volume-> Change..- > Select a new host-> apply
  5. the cold standby starts and the system scans directly to the target block.
  6. Mount partition

it is worth mentioning that after entering the system, because of the DM-MP method used, Linux will scan the storage block device directly through the built-in Devicemapper , without the need to install additional multichannel services.

the reason why I did not choose grouping here is that grouping is suitable for cluster and dual-computer use. Each host occupies a partition authorization, but at present, it is authorized exclusively. I am afraid it will be an unnecessary accident to change it hastily. In addition, when the cold backup conditions are met, once the old server crashes, the cold backup boot sequence is the above steps, which is actually very time-consuming and can not meet the requirements of the national standard (this standard is said to be non-compliant).

I also consulted with the customer why we didn't do the dual machine hot. What the customer meant was that they didn't understand this at the beginning, and the bidding company entrusted when the project was tendered did not list this technical requirement. During the bid evaluation, the other party only listed the quotation executed by the stand-alone machine in the contract, and the backup plan was cold standby. After the enterprise won the bid and deployed, the customer negotiated with it to buy a dual-computer hot backup solution, and the other party quoted a double price, but the technical implementation was only a HA of the same license on the same service.

at that time, the third company offered dual-computer hot backup price was 1/3 of the current business system.

. This is definitely a trick.

Previous: How to get the width and height of a picture when uploading a picture?

Next: Is the performance of mysql: limit much worse than that of between or in?

css mysql arrays josn react html typescript webpack npm sass R objective-c .net sql-server jquery python-3.x angularjs django angular excel regex iphone ajax linux xml pandas vba spring database wordpress string wpf xcode windows bash postgresql oracle multithreading eclipse list firebase algorithm macos forms image scala visual-studio azure bootstrap spring-boot react-native python-2.7 docker performance function winforms matlab powershell apache dataframe api sqlite numpy rest shell selenium flutter dart maven loops qt swing android-studio csv express file class tensorflow sorting codeigniter perl
MySQL Query : SELECT * FROM `codeshelper`.`v9_news` WHERE status=99 AND catid='6' ORDER BY rand() LIMIT 5
MySQL Error : Disk full (/tmp/#sql-temptable-64f5-1e87d2b-e8f.MAI); waiting for someone to free some space... (errno: 28 "No space left on device")
MySQL Errno : 1021
Message : Disk full (/tmp/#sql-temptable-64f5-1e87d2b-e8f.MAI); waiting for someone to free some space... (errno: 28 "No space left on device")
Need Help?