Recently, I noticed that I cannot run any snapins that were associated with the host. I also noticed the FOGScheduler was not working. After investigating, I found several issues.

FOG Version:
Upgraded from 1.15.10 → <several 1.6-beta in between> → currently running 1.6.0-beta.2297

Setup:

• FOG Server: 172.28.1.80
• Storage Node: 172.28.1.89

1) Replicator: falsely reports image files as missing at first, then immediately syncs.

[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.fixed_size_partitions(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.mbr(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.minimum.partitions(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.original.fstypes(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.original.swapuuids(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.partitions(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1.shrunken.partitions(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1p1.img(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1p2.img(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1p3.img(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist d1p4.img(storage 1)
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.fixed_size_partitions on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.fixed_size_partitions file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.mbr on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.mbr file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.minimum.partitions on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.minimum.partitions file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.original.fstypes on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.original.fstypes file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.original.swapuuids on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.original.swapuuids file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.partitions on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.partitions file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1.shrunken.partitions on storage 1
[04-07-26 9:28:10 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1.shrunken.partitions file to storage 1
[04-07-26 9:28:10 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1p1.img on storage 1
[04-07-26 9:28:11 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1p1.img file to storage 1
[04-07-26 9:28:11 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1p2.img on storage 1
[04-07-26 9:28:11 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1p2.img file to storage 1
[04-07-26 9:28:11 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1p3.img on storage 1
[04-07-26 9:28:11 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1p3.img file to storage 1
[04-07-26 9:28:11 am]   # peruswin-audit-1.2: File does not exist on master node, deleting /images/peruswin-audit-1.2/d1p4.img on storage 1
[04-07-26 9:28:11 am]  | peruswin-audit-1.2: No need to sync /images/peruswin-audit-1.2/d1p4.img file to storage 1
[04-07-26 9:28:11 am]  | CMD: lftp -e 'set xfer:log 1; set xfer:log-file "/opt/fog/log/fogreplicator.peruswin-audit-1.2.transfer.storage 1.log";set ftp:list-options -a;set net:max-retries 10;set net:timeout 30; mirror -c --parallel=20 -R --ignore-time -vvv --exclude ".srvprivate" "/images/peruswin-audit-1.2" "/images/peruswin-audit-1.2";exit' -u fogproject,[redacted] 172.28.1.89
[04-07-26 9:28:11 am]  * Started sync for Image peruswin-audit-1.2 - Resource id #1583
[04-07-26 9:28:11 am]  | Sync finished - Resource id #602

Observed:

• Files exist on both server and storage under /images/<image>
• Verified with find and lftp
• Image deploys successfully to clients

2) SSH spam between nodes

Apr 07 09:42:13 fog sshd[2483177]: error: kex_exchange_identification: Connection closed by remote host
Apr 07 09:42:13 fog sshd[2483177]: Connection closed by 172.28.1.89 port 55330
Apr 07 09:42:13 fog sshd[2483178]: error: kex_exchange_identification: Connection closed by remote host
Apr 07 09:42:13 fog sshd[2483178]: Connection closed by 172.28.1.89 port 55336
Apr 07 09:42:14 fog sshd[2483179]: error: kex_exchange_identification: Connection closed by remote host
Apr 07 09:42:14 fog sshd[2483179]: Connection closed by 172.28.1.80 port 34766
Apr 07 09:42:14 fog sshd[2483180]: error: kex_exchange_identification: Connection closed by remote host
Apr 07 09:42:14 fog sshd[2483180]: Connection closed by 172.28.1.80 port 34768

Observed:

• Happens multiple times per second
• Seen on both server and storage

Fix / Isolation:

• Stopping FOGMulticastManager stops the SSH spam
• Starting it again reproduces the issue

3) FOGMulticastManager creates broken PHP session files (storage node)

session_start(): open(... Permission denied)

Observed:

• /var/lib/php/sessions directory is correct:

  drwx-wx-wt root:www-data
  

• Session files are created as:

  -rw------- 1 root root ...
  

• Apache/PHP-FPM runs as www-data → cannot access them

Isolation:

• Stop:

  systemctl stop FOGScheduler FOGMulticastManager
  

• Delete sessions:

  find /var/lib/php/sessions -type f -name 'sess_*' -delete
  

• Errors stop

• Start only:

  systemctl start FOGMulticastManager
  

• Errors immediately return

4) Power Management warnings

Undefined array key "pmAction"

Observed:

• Many hosts have no row in powerManagement table

Fix:

• Disabling Power Management in FOG settings stops the warnings

5) Scheduler tasks do not run

Observed:

• Scheduled tasks do not execute unless FOGScheduler is restarted
• After restart, tasks run, but later scheduler stalls again
• New tasks are not picked up

6) Snapins do not execute

Observed:

• Snapins can be assigned
• Execution on a single host associated with the snapins fails
• Snapin runs on a group

7) High CPU usage (PHP)

php (root) ~100% CPU

Observed:

• High CPU usage on both server and storage
• Drops when stopping FOGScheduler

Additional notes

• Manual SSH from server → storage using fogproject works
• FTP (lftp) can list image files correctly
• Installer has been re-run on both nodes after update