Resolved Issues in Release 23.5.30-PCE
PCE setup does not work on RHEL 9.x in FIPS mode (E-119668)
This release resolves an incompatibility with the PCE and RHEL 9.x in FIPS mode which caused the PCE to not start properly.
Errors in Flow Analytics (E-118558)
Flows in Illumination or the traffic database summary were not appearing, and the traffic database size summary was being shown as zero on some PCE clusters.
Last updated policy timestamp for C-VENs reflects Kubernetes Workload policy changes (E-118372)
The last updated policy timestamp on C-VENs now updates after a C-VEN successfully updates the policy for its pods.
Navigation error while navigating to Authentication Settings > SAML: Not Found (E-118183)
In PCEs running 22.5.32, sometimes going to Authentication Settings > SAML resulted in the attempted navigation being cancelled, and a "
Navigation error details
" popup appearing.PCE is sending partial IPP instructions (E-117863)
PCE was sending partial IPP instructions to Kubelink, which resulted in incorrect policy in the destination cluster.
Policy generator throwing an error when saving rules (E-117499)
When users tried to save the rule with custom iptables rules, the Policy generator was throwing an "Unexpected input validation error".
Missing app-tiers label on pod using annotation (E-117004)
In non-CLAS (legacy) container clusters, when applying Illumio labels through Kubernetes annotations, a label key containing a dash (-) is not properly assigned to Container Workloads. For example, a pod annotation of
annotation.com.illumio.app-tiers
with a label value of AT_A is not created with label typeApp-Tiers
nor the labelAT_A
. This issue is now resolved for new Container Workloads created on this release. However, upgrading the PCE to this release does not fix existing Container Workloads that have labels containing a dash character. To fix such existing Container Workloads, you can edit the Container Workload Profile to add another possible value for the dash-containing label. After saving this edit, existing Container Workloads get re-labelled correctly to their assigned annotation values.NEN 2.6.20 is stuck in "ACL generation pending" (E-116805)
In a configuration with a 2.6.20 NEN paired with a supercluster member on PCE Version 22.5.32-12, running "Generate ACLs" never completed, and only showed the "
ACL Generation Pending
" message without ever producing an ACL.CLAS - Rules are not created for Kubernetes Workloads and VIPs (E-116721)
In CLAS-enabled deployments, rules created between a Kubernetes Workload and a VIP (from a virtual server, for example a F5 Virtual Server) are not created even after provisioning. These rules fail to appear in the PCE Web Console. This issue is resolved. The new runtime environment variable
clas_workloads_ipset_only_changes_enabled
must be set tofalse
in the PCEruntime_env.yml
file (underagent_service
:) for the PCE to correctly send Virtual Server instructions to Kubernetes Workloads.Header manipulation issue fixed (E-116114)
Appropriate validation for host header was added to avoid any host header manipulation.
HTTP 500 error from Kubernetes Workloads filter (E-115416)
After navigating to Workloads > Kubernetes Workloads, then setting the Filter to Category NO LABEL, Type == NO APPLICATION LABELS, after pressing Enter, the action fails with a Navigation has been cancelled due to an error message appears.
Container cluster reporting "Virtual service is still active on a workload" after upgrading to "clusterMode: migrateLegacyToClas" (E-114727)
After a non-CLAS (legacy) deployment was upgraded to CLAS mode, existing container clusters running multiple ClusterIP virtual services each went into an Error Status, with each cluster detail page also displaying a "
Virtual service is still active on a workload
" message.report_monitor and traffic_query services flapping on coordinator replica node after OS upgrade (E-113024)
On DX configurations, adding a new CC (Citus Coordinator) node or a new CW (Citus Worker) node to the cluster sometimes caused flapping of some services, such as report_monitor or traffic_query. This flapping occurred because IP restrictions on some current nodes of the cluster did not account for the new node IP addresses.
External users with multiple scopes reporting PCE slowness (E-109314)
External users with many scopes in their RBAC permission have been reporting PCE UI slowness, especially when browsing the VENs tab and querying traffic.