Stacking cable replacement

I have some stacks built with M4300 switches. Stack connections are made using 10GBASE-T ports with ring topology. Each stack consists 3 units, 2x M4300-28G and 1x M4300-28G-PoE+. Recently I've found multiple messages in device logs, here are examples:

<13> May 16 12:48:10 PD4-1 TRAPMGR[spmTask]: traputil.c(795) 1441345 %% Stack-port link up: Index: 201 Unit: 2 Tag: 2/0/26
<13> May 16 12:48:10 PD4-1 TRAPMGR[spmTask]: traputil.c(795) 1441344 %% Stack-port link up: Index: 300 Unit: 3 Tag: 3/0/25
<13> May 16 12:48:10 PD4-1 TRAPMGR[spmTask]: traputil.c(795) 1441341 %% Stack-port link down: Index: 201 Unit: 2 Tag: 2/0/26
<13> May 16 12:48:10 PD4-1 TRAPMGR[spmTask]: traputil.c(795) 1441339 %% Stack-port link up: Index: 201 Unit: 2 Tag: 2/0/26

And even worse, on one stack there are many messages like that:

<10> May 16 14:02:05 PD4-1 STACKING[spmTask]: spm.c(1419) 1442390 %% Multiple consecutive errors detected on stack-port 2/0/26 (oldRxErrors = 190996 currentRxErrors = 190997 oldTxErrors = 0 currentTxErrors = 0). Please fix the errors before proceeding. The errors would likely be resolved by changing stack cables

As far i I know, some (or all) stacking cables needs to be replaced. My question is:

It is possible to do this without interrupting normal work? And Is there any special procedure, or just replace one by one, and how long wait between?

Ring topology is redundant, so it seems to be possible replace cables just one by one, but I want to be sure.