cscott/ask-annie
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add chapters for session 946343310 (Haiku)

Browse Source
This commit is contained in:
Conan Scott
2026-03-24 10:07:49 +00:00
parent a8532f7d5c
commit df9885000c
1 changed files with 12 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
chapters/946343310.json Normal file
View File

@@ -0,0 +1,12 @@
[
{"timestamp": "2:44", "title": "Red Hat 8 to Red Hat 9 Migration Strategy", "summary": "Three approaches for migrating enterprise cluster: (1) Install new servers and add to cluster with database ID and Doctor implications; (2) Build as separate Doctor system to preserve configuration; (3) Clean build against new database with import/export, losing audit logs and tracking tables. New servers cannot be simply added due to ID conflicts and Doctor replication issues. Preferred approach depends on whether Doctor systems exist and acceptable data loss."},
{"timestamp": "10:31", "title": "Migration Practical Considerations", "summary": "Migration success depends on storage mounts, DNS resolution, LDAP/SSO IP whitelisting, antivirus rules, and mail notification settings. Edges should be rebuilt completely — no migration of edge configuration is recommended. Only XML import/export works cleanly between old and new systems."},
{"timestamp": "16:45", "title": "Zero Downtime Upgrade vs OS Migration Differences", "summary": "Zero Downtime Upgrade (ZGU) keeps one database with multiple replicated schemas, switching servers dynamically. OS migration uses separate databases and Doctor model, causing service disruption. ZGU does not change OS; OS migration must. Future ZGU may support OS migration but currently is not recommended for this use case."},
{"timestamp": "24:36", "title": "SMB Connector Protocol and Use Cases", "summary": "SMB connector is used for transferring files from on-premises Samba servers to cloud ST instances. Port is configurable per transfer site. SMB establishes one connection per file with mounting overhead — not suitable for high-volume transfers (>100 files/hour). CFT or similar tools recommended for large-scale ground-to-cloud file movement."},
{"timestamp": "31:42", "title": "SSH Authentication Timeout and Network Diagnostics", "summary": "SSH authentication taking ~5 minutes with public key signing failures is likely caused by firewall or load balancer idleness timeout, not ST configuration. ST uses Maverick SSH libraries (stricter RFC compliance) versus OpenSSH (more flexible). Typical authentication should complete in <30 seconds. Manual SSH client test from edge server confirms if issue is with partner's noncompliant SSH server."},
{"timestamp": "39:55", "title": "SFTP Connection Timeouts and Server Compatibility", "summary": "Five-minute timeout pattern suggests external firewall/load balancer closing connections, not ST. Maverick libraries may expose noncompliant servers that OpenSSH accepts. Test direct connection from edge to partner; if slowness persists, investigate partner's SSH server type and compatibility."},
{"timestamp": "40:00", "title": "Multi-Destination File Routing Setup", "summary": "When routing files to both external SSH and internal network path via folder monitor fails with 'path not found': verify path accessibility from ST server OS level (not desktop), check for typos and case sensitivity, use FQDN or IP instead of short hostnames, and verify Windows share accessibility from Linux application context."},
{"timestamp": "42:03", "title": "Network Path Resolution Issues (Linux to Windows)", "summary": "ST runs as Linux application accessing Windows shares. Case sensitivity, short hostname resolution, and DNS configuration can cause UNC path failures. Verify path from ST server command line, test with FQDN or IP address, and ensure Windows share ports are open to ST servers."},
{"timestamp": "46:50", "title": "Account Maintenance Multiple Threshold Configuration", "summary": "Only one account maintenance application instance can run per system. Multiple thresholds for key expiration warnings must be configured within a single application using comma-separated day values in custom account fields (e.g., 30,60,90). Workaround to disable a threshold: set impossibly high value like 9999."},
{"timestamp": "52:18", "title": "Key Expiration Notification Timing", "summary": "Account maintenance sends notifications on configured days before expiration. If server is down on notification day, notification does not retroactively send. Use staggered values with buffer days (e.g., 30, 55, 60) to ensure at least one notification triggers. Field format for multiple values is not well-documented."}
]