[Bugs] [Bug 1593826] Glusterfs being killed due to out of memory situation

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1593826



--- Comment #11 from hari gowtham <hgowtham at redhat.com> ---
The protocol clients seem to be consuming a lot of resource:

Client 0:
[protocol/client.gDBv1-client-0 - usage-type gf_common_mt_memdup memusage]      
size=151041876                                                                  
num_allocs=12586878                                                             
max_size=151042352                                                              
max_num_allocs=12587210                                                         
total_allocs=159274593 

[protocol/client.gDBv1-client-0 - usage-type gf_common_mt_char memusage]        
size=311525526                                                                  
num_allocs=12586886                                                             
max_size=311534067                                                              
max_num_allocs=12587217                                                         
total_allocs=226411352 

Client 2:
[protocol/client.gDBv1-client-2 - usage-type gf_common_mt_memdup memusage]      
size=151042232                                                                  
num_allocs=12586936                                                             
max_size=151042348                                                              
max_num_allocs=12587210                                                         
total_allocs=159380212

[protocol/client.gDBv1-client-2 - usage-type gf_common_mt_char memusage]        
size=311526832                                                                  
num_allocs=12586946                                                             
max_size=311533894                                                              
max_num_allocs=12587219                                                         
total_allocs=226444611 

and so on for the other clients too.
As seen its high with gf_common_mt_memdup and gf_common_mt_char 

To verify this, the statedumps at interval from Edward will be helpful.

It would be great if someone from protocol team can take a look.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.