Tag
13 articles
Set up a hardened chrooted SFTP server on CentOS Stream 10 with OpenSSH internal-sftp, key-only auth, firewalld and SELinux - and fix the connection closed and permission denied errors that follow.
12 Sep 20267 min
Build a production NFS server on Ubuntu 26.04 LTS - exports, NFSv4 firewall rules, client mounts - then fix access denied, RPC not registered, stale file handles and nobody:nogroup ownership.
12 Sep 20268 min
Broadcom documents the bugs that took down somebody's cluster. Five real issues from the latest ESXi release notes - the exact error, the root cause and the fix, with the war stories behind them.
12 Sep 202612 min
Install Failed on one host or the whole cluster tells you different things. Read esxupdate.log, clear the real blocker and recover a half-installed transport node without a rebuild.
12 Sep 20268 min
BGP stuck in Active, or Established with no prefixes? Two very different problems. A structured walk from Tier-1 advertisement through Tier-0 redistribution to the physical router.
12 Sep 20268 min
A degraded NSX Manager cluster leaves the data plane running but configuration at risk. Work through cluster status, disk, time, certificates and a clean node rejoin.
12 Sep 20268 min
A DFW rule that looks correct in the UI but does not match in the data path. Use vsipioctl, Traceflow and filter captures to find the rule that really handled the packet.
12 Sep 20269 min
Overlay traffic between hosts stops but NSX looks green. Here is how to read tunnel state, prove TEP reachability at 1600 bytes and find the MTU or BFD fault behind it.
12 Sep 202610 min
APD, PDL and plain latency produce similar symptoms and need completely different responses. Here is how to classify the condition from the logs and recover correctly.
11 Sep 20268 min
Consolidation warnings, locked delta disks and orphaned VMDK chains all come from the same mechanics. Learn the disk chain, the safe recovery order, and the alarms that stop it happening again.
11 Sep 20267 min
Most 'HA agent unreachable' errors are DNS, firewall or management-network faults rather than HA itself. This fault tree takes you to the root cause in seven checks.
10 Sep 20268 min
The appliance pings and accepts SSH, but the vSphere Client returns 503. Work through disk space, certificates, time skew, directory service and database in the right order.
9 Sep 20267 min
A purple diagnostic screen is the VMkernel telling you it hit an unrecoverable condition. Here is how to read the panic banner, extract the core dump, and identify the faulting driver or hardware component.
8 Sep 20267 min
