LDAP monitor with error : fork() failed: Cannot allocate memory

[Sarovani]

Hi Experts ,

We have LDAP VIP/pool which has custom LDAP health check monitor associated with it . But the health check monitor has marked the pool member as down with the error : fork() failed: Cannot allocate memory.

If we change the monitor with default tcp , the pool will be up .and also we have F5 in cluster .So this issue is there only on the Active F5 , On standby f5 the pool is up with same LDAP custom montor .

Dec 6 17:33:11 f5-lan-primary notice mcpd[6358]: 01070638:5: Pool /Common/pool_ldap_prd member /Common/192.168.1.11:636 monitor status down. [ /Common/m_ldap_prd: down; last error: /Common/m_ldaps_prd: fork() failed: Cannot allocate memory @2022/12/06 13:49:09. ] [ was up for 0hr:1min:1sec ]

Dec 6 17:33:11 f5-lan-primary notice mcpd[6358]: 01070638:5: Pool /Common/pool_ldap_prd member /Common/192.168.1.12:636 monitor status down. [ /Common/m_ldap_prd: down; last error: /Common/m_ldaps_prd: fork() failed: Cannot allocate memory @2022/12/06 15:57:29. ] [ was up for 0hr:1min:1sec ]

Can you please check and advice , if you are familiar with this error message ? Please note there is no issue with the memory on the F5 .Looks normal .

[Mohamed_Ahmed_Kansoh]

Hi @Sarovani , 
Maybe if your Restart System services ( tmsh restart sys services ) or (bigstart restart) , it will work normally with you. 
But I do not prefer this solution as it will impact your on air services and you need a maintenance Window to do it. 

> I recommend to contact with F5 support team for this issue after referring to this

KB : https://support.f5.com/csp/article/K50030452
it recommends to consult with F5 support team for ( memory allocation faild issues ) 

Regards

[Mike757]

Hi Sarovani,

There seems to be nothing wrong with your monitor configuration, since the other F5 box shows no issues and the pool members are up.

But Mohamed ir quite right, you should restart services (or maybe even do a full_box_reboot). I would follow this general procedure:

- schedule a maintenance window (30 min should be enough)
- confirm DSC is in-sync
- force a failover
- check everything is up and running in the newly active system
- execute "full_box_reboot" on the standby system
- check status after reboot
- if all ok (LDAP monitor ok on those pool members), optionally failover again

And by the way... what version are you running? A memory leak bug may already have been corrected in newer images.

/Mike

[Sarovani]

Hi @Mike757 @Mohamed_Ahmed_Kansoh 

Thank you for the reply , We have tried belo steps 

On f5-lan-primary : Did a failover and made f5-lan-secondry as active .

1.Restarted the F5 from CLI - No luck 

2.Restarted the F5 from VM console - issue got resolved .

But strange , why the first reboot from F5 cli did not fix the issue . We hve cse open with F5 support waiting for RCA .

[Mohamed_Ahmed_Kansoh]

Hi @Sarovani , 

Restarting services differs from Rebooting the Full VM which you restart F5 Kernel and Whole Daemons. 

But I would like to Know the Reason from F5 support team. 
so update us when you know about it.