Problem Description:

VCSA running well suddenly cannot be accessed, and the web page reports an error:
503 service unavailable (failed to connect to endpoint: [[n7vmacore4http16localservicespec: 0x00005649e04415e0] _servernamespace =/_isredirect = false _pipename =/var/run/VMware/vpxd webserver pipe)

Cause analysis:

https://kb-uat.vmware.com/s/article/67818?lang=zh_cn

The reasons I often encounter are:

vCenter server appliance (VCSA) high virtual disk space (76563) check the expiration date of the certificate on vCenter server (79248/82332) vSphere vxpd service of VCSA is abnormal and does not operate normally;

---

Solution:

1、 VCenter server appliance (VCSA) high virtual disk space (76563)

https://kb-uat.vmware.com/s/article/76563?lang=zh_cn

log in to the vCenter server appliance as root through SSH or through the vCenter virtual machine console to find out which partitions are full;

df -h

Run the following command to find partitions that are 85% full or higher:

df -h |awk '0+$5 >= 85 {print}'

Find large log files and delete them if they are useless. Note: *. Tgz files are archived log package files and can be removed.

find /storage/log -type f -size +100M

2、 Check the expiration date of the certificate on the vCenter server (79248/82332)

https://kb.vmware.com/s/article/79248?lang=zh_ cn
https://kb-uat.vmware.com/s/article/82332

The expiration of this certificate is often encountered on vcsa6.0, because the STS certificate of vcsa6.0 generally expires in 2 years. It is very difficult to hang up as soon as it expires. Encountered two or three times. According to the official KB, I failed to handle it successfully once. I don’t understand why ~. Finally, the toss is to redeploy the VCSA, which is faster and easier (provided that there are fewer managed hosts on your VCSA, and it’s not troublesome to add it again)
however, this problem has not been encountered in vcsa6.5 and above. If you encounter it, please refer to the official KB.

3、 VSphere vxpd service of VCSA is abnormal and is not running normally

In the past, I deployed VCSA myself as “vCenter server with an embedded platform services controller”, and then encountered 503 errors. VCSA also looked at the space utilization and vSphere client services, but vSphere vxpd services couldn’t work. After a closer look, I found that the VCSA architecture was “vCenter server with an external platform services controller”, There is also a separate platform services controller virtual machine. It is suspected that there is a problem with the PSC, but the root password is not known, so it can only be forcibly restarted. Due to abnormal shutdown, the startup enters the emergency rescue mode,
prompt failed to start system check on/dev/disk/by partuuid/* * *
backhand is one

fsck /dev/disk/by-partuuid/***

Restart it again, start it normally, return to VCSA and restart all services

service-control --stop --all
service-control --start --all

So far, VCSA has finally been accessed normally.