No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>

Reminder

To have a better experience, please upgrade your IE browser.

upgrade

How to Troubleshoot the Fault that the Cluster NFS Service Becomes Abnormal

Publication Date:  2012-07-18 Views:  33 Downloads:  0
Issue Description
The NFS service becomes abnormal at possibilities after the file system is shared to the specified IP addresses.
Symptom:
N8000_01: # hagrp –state |grep NFS
NFS                  State              N8000_01 |PARTIAL|STARTING|
NFS                  State              N8000_02 |PARTIAL|STARTING|
NFS                  State              N8000_03 |PARTIAL|
NFS                  State              N8000_04 |PARTIAL|
If the preceding symptom occurs, the client access to the sharing space of the N8000 cluster is affected.
Alarm Information
None
Handling Process

Note: The following operations are conducted in the CLI.
Solution 1: Disable the DNS service (recommended).
Step 1 Log in to the IP address of the clustered NAS engine console using the master account and disable the DNS service.
N8000> network dns disable
Please enable/disable DNS in nsswitch settings for required services.
N8000 dns Success V-288-1033 successful
If successful is displayed, the DNS service is successfully disabled.
Step 2 Check whether the exportfs process exists.
N8000_01:/hs # ps -eaf | grep exportfs
root     23067 30385  0 15:41 pts/4    00:00:00 grep exportfs
The result shows that the process does not exist.
Step 3 Check whether the share resource restores to normal.
N8000_01: # hares –state |grep share
SHARE_103                  State              N8000_01 ONLINE
SHARE_103                  State              N8000_02 ONLINE
SHARE_103                  State              N8000_03 ONLINE
SHARE_103                  State              N8000_04 ONLINE
SHARE_103                  State              N8000_05 ONLINE
SHARE_103                  State              N8000_06 ONLINE
SHARE_103                  State              N8000_07 ONLINE
SHARE_103                  State              N8000_08 ONLINE
SHARE_103                  State              N8000_09 ONLINE
SHARE_103                  State              N8000_10 ONLINE
SHARE_103                  State              N8000_11 ONLINE
SHARE_103                  State              N8000_12 ONLINE
 If ONLINE is displayed, the recovery is normal. Otherwise, go to step 4.
Step 4  Implement the automatic recovery as user master.
N8000>support services autofix                                                                   
Attempting to fix service faults..........done
 If done is displayed, no further action is required.
Step 5 Check whether the NFS service restores to normal.
N8000_01: # hagrp –state |grep NFS
NFS                  State              N8000_01 ONLINE
NFS                  State              N8000_02 ONLINE
NFS                  State              N8000_03 ONLINE
NFS                  State              N8000_04 ONLINE
If ONLINE is displayed, the NFS service restores to normal.
Solution 2: Recover the DNS service.
Step 1 Check whether the DNS service is enabled.
N8000> network dns show
DNS Status      : Enabled
domain          : hs.com
nameserver      : 172.24.82.1
If Enabled is displayed, the DNS service is enabled.
Refer to the related DNS server documents to locate and troubleshoot faults.
Step 2 Ping the recovered DNS server.
N8000> network ping 172..24.82.1                                                                                     
PING 172.24.82.1 (172.24.82.1) 56(84) bytes of data.
64 bytes from 172.24.82.81: icmp_seq=1 ttl=64 time=0.165 ms
64 bytes from 172.24.82.81: icmp_seq=2 ttl=64 time=0.112 ms
64 bytes from 172.24.82.81: icmp_seq=3 ttl=64 time=0.097 ms
64 bytes from 172.24.82.81: icmp_seq=4 ttl=64 time=0.103 ms
64 bytes from 172.24.82.81: icmp_seq=5 ttl=64 time=0.150 ms
--- 129.62.21.199 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 3999ms
rtt min/avg/max/mdev = 0.097/0.128/0.165/0.028 ms
The result shows that the DNS service is usable.
Step 3 After the cluster shares the file system to a specified IP address, no NFS anomaly occurs.
N8000_01: # hagrp –state |grep NFS
NFS                  State              N8000_01 ONLINE
NFS                  State              N8000_02 ONLINE
NFS                  State              N8000_03 ONLINE
NFS                  State              N8000_04 ONLINE

Root Cause
Check whether a file system hangs and then check whether the DNS service is correctly configured.
1. Log in to the cluster as the support user using SSH. The IP address is the IP address of the console. The default password is symantec. Then do as follows:
Step 1 Check whether the file system is identified as fsck.
N8000_01: echo "8192B.p S" | fsdb -t vxfs /dev/vx/rdsk/sfsdg/fs_name  |grep flag
flags 4000 mod 0 clean 3c
If the last value of flag is not 1, the file system is normal and the fsck command does not need to be used.
Step 2 Check whether the bottom-layer share resource is correct.
N8000_01:/hs # hares –state |grep share
share_103                   State                   N8000_01 FAULTED
share_103                   State                   N8000_02 FAULTED
share_103                   State                   N8000_03 FAULTED
share_103                   State                   N8000_04 FAULTED
The share resource becomes abnormal on the node.
Step 3 Check whether the file system hangs on the node. Run the following command to determine the master node. The result shows that node 1 is the master node.
N8000_01: # vxdctl –cmode
mode: enabled: cluster active – MASTER
master: N8000_01
Step 4 Run the following command on the master node:
N8000_01:/hs # ps -eaf | grep exportfs
root     22702 22701  0 15:31 ?        00:00:00 exportfs -u *:/vx/SamePlatData
root     23064 30384  0 15:31 pts/4    00:00:00 grep exportfs
In normal cases, when the share add + client command is executed successfully, the exportfs process does not exist on the node. If the exportfs process exists, the file system hangs when configuring the share.
Step 5 Log in as user support and run the su – master command to switch to user master. Then run the following command to check the status of the cluster DNS service:
N8000> network dns show
DNS Status : Enabled
domain : hs.com
nameserver 172.24.82.1
Step 6 Ping the DNS server.
N8000> network ping 172..24.82.1
PING 172.24.82.1 (172.24.82.1) 56(84) bytes of data.
From 172.24.82.81: icmp_seq=2 Destination Host Unreachable
From 172.24.82.81 icmp_seq=2 Destination Host Unreachable
From 172.24.82.81 icmp_seq=3 Destination Host Unreachable
From 172.24.82.81 icmp_seq=4 Destination Host Unreachable
From 172.24.82.81 icmp_seq=5 Destination Host Unreachable
--- 129.62.21.9 ping statistics ---
5 packets transmitted, 0 received, +5 errors, 100% packet loss, time 4000ms
pipe 3
N8000  ERROR V-288-825 Could not connect
The result shows that the DNS service is enabled but the DNS server cannot be pinged through.
2. During the share creation, the created file system is exported. If the DNS service is enabled, the NFS server looks up the client on the DNS server. If the domain name does not exist, the file system continuously implements the exportfs operation. However, the file system cannot be properly exported. As a result, the file system hangs on the node and the NFS service status becomes abnormal. The fault occurs at possibilities.
Suggestions
By default, the DNS service is enabled after the N8000 is installed. If you do not want to use the DNS service, manually disable it.

END