From 8dd42a478a39e70e752eb2219c3fd9e655d79fcf Mon Sep 17 00:00:00 2001
From: Konrad Bucheli <konrad.bucheli@psi.ch>
Date: Tue, 10 Oct 2023 11:39:26 +0200
Subject: [PATCH] how to handle PCIe bus errors

---
 _toc.yml                                      |  1 +
 admin-guide/troubleshooting/pcie_bus_error.md | 18 ++++++++++++++++++
 2 files changed, 19 insertions(+)
 create mode 100644 admin-guide/troubleshooting/pcie_bus_error.md

diff --git a/_toc.yml b/_toc.yml
index 6ae68a26..d21ad5e1 100644
--- a/_toc.yml
+++ b/_toc.yml
@@ -103,6 +103,7 @@ chapters:
           - file: admin-guide/troubleshooting/boot
           - file: admin-guide/troubleshooting/kerberos
           - file: admin-guide/troubleshooting/sssd
+          - file: admin-guide/troubleshooting/pcie_bus_error
       - file: admin-guide/order-vm
 
   - file: infrastructure-guide/index
diff --git a/admin-guide/troubleshooting/pcie_bus_error.md b/admin-guide/troubleshooting/pcie_bus_error.md
new file mode 100644
index 00000000..140c6df8
--- /dev/null
+++ b/admin-guide/troubleshooting/pcie_bus_error.md
@@ -0,0 +1,18 @@
+# PCIe Bus Error
+
+When there are PCI Express bus errors like
+```
+Oct 05 11:26:19 pc16209.psi.ch kernel: pcieport 10000:e0:06.0: AER:   TLP Header:  34000000 e1000010 89148914 00000000
+Oct 05 11:26:19 pc16209.psi.ch kernel: pcieport 10000:e0:06.0: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID)
+Oct 05 11:26:19 pc16209.psi.ch kernel: pcieport 10000:e0:06.0:   device [8086:464d] error status/mask=00100000/00010000
+Oct 05 11:26:19 pc16209.psi.ch kernel: pcieport 10000:e0:06.0:      [20] UnsupReq          (First)
+```
+you might try with disabling **Active State Power Management** (ASPM) in the kernel.
+
+To do so set in Hiera
+
+```
+base::enable_pcie_aspm: false
+```
+
+the apply it with `puppet agent -t` and reboot.