search

hautreux / auks

Kerberos credential support for batch environments
Other
20 stars 18 forks source link

Compute node reports: `auks api : request processing failed` #67

Open phuelsdunk opened 2 years ago

phuelsdunk commented 2 years ago

Hello,

I have troubles debugging the auks -p call on a compute node node2:

[root@node2 auks]# auks -p
Auks API request failed : auks api : request processing failed

On this node, aukspriv is running and it adds the correct default principal:

[root@node2 auks]# klist
Ticket cache: FILE:/tmp/krb5cc_0
Default principal: host/node2.lan@REALM.A

Valid starting       Expires              Service principal
12.08.2022 09:45:11  12.08.2022 19:45:11  krbtgt/REALM.A@REALM.A
        renew until 19.08.2022 09:45:11

On the login and management node node1, aukspriv, auksd & auksdrenewer are running, here the auks call works successfully:

[root@node1 ~]# klist
Ticket cache: FILE:/tmp/krb5cc_0
Default principal: host/node1.lan@REALM.A

Valid starting       Expires              Service principal
12.08.2022 09:55:13  12.08.2022 19:55:13  krbtgt/REALM.A@REALM.A
        renew until 19.08.2022 09:55:13
[root@node1 ~]# auks -p
Auks API request succeed
[root@node1 ~]# klist
Ticket cache: FILE:/tmp/krb5cc_0
Default principal: host/node1.lan@REALM.A

Valid starting       Expires              Service principal
12.08.2022 09:55:13  12.08.2022 19:55:13  krbtgt/REALM.A@REALM.A
        renew until 19.08.2022 09:55:13
12.08.2022 09:55:19  12.08.2022 19:55:13  host/node1.lan@REALM.A
        renew until 19.08.2022 09:55:13

Any ideas how to debug this further?

I am using these configuration files:

fihuer commented 2 years ago

Hi, This sounds like a Kerberos-related issue. Could you try to reproduce with KRB5_TRACE=/dev/stderr in your environnment ? Adding the auksd log file here would help too.

btw, we are currently transferring the ownership of auks to another repo (https://github.com/cea-hpc/auks), could you re-open a Discussion/Issue over there ?