Large number of warnings generated - cea_disk_large_object_handler

Products

• ISG Proxy

Issue/Introduction

ProxySG Appliance Event 48025

2023-10-20 07:44:53-00:00UTC  "'https://xxxxxxxxxxxxxxxxxxxx/?url=httpxxxxxxxxxxxxxxxx' web object, which is 427830099 bytes long. (Flags 700004, 2000)"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:84

2023-10-20 07:44:53-00:00UTC  "'CFS/CB3/flop/database.app' persistent object, which is 37936728 bytes long. (Flags 100012, 800050)"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:84

2023-10-20 07:44:53-00:00UTC  "- Found 400 large web objects, 0 are in use, consuming 77,755 MB, some of which are:"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:364

2023-10-20 07:44:53-00:00UTC  "- Found 10 large persistent objects, 0 are in use, consuming 3,270 MB, some of which are:"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:353

2023-10-20 07:44:53-00:00UTC  "Disk 2 contains 56231 web object, 3848 system objects, 54003 small objects, 6076 large objects, 19/36 newly-created/in-use objects, 935 deletion objects, 0 terminating objects and 0 objects with IO errors."  0 48025:1 Mailed cea_disk_large_object_handler.cpp:342

2023-10-20 07:44:53-00:00UTC  "'is_categorization/failed.tar' persistent object, which is 2007454488 bytes long. (Flags 500012, 800050)"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:84

2023-10-20 07:44:53-00:00UTC  "'is_web_app/active.bin' persistent object, which is 42804353 bytes long. (Flags 100012, 800050)"  0 48025:1 Mailed cea_disk_large_object_handler.cpp:84

 

ProxySG Appliance Event 4801E

2023-10-20 07:44:53-00:00UTC  "Disk 2 has only 4% free space (642090 blocks free out of 15596748) for 59725 objects. Seconds until next print message: 300"  0 4801E:1 Mailed ced.cpp:1002

Cause

The "cea_disk_large_object_handler" messages are all related to huge files being downloaded (antivirus scanned thanks to ICAP) via proxy and filling up the system storage available space.

The issue happens when the SG-VA runs out of disk space and when this happens the SG-VA will be reinitialized and hence subsequent loss of disk space. This can happen due to ICAP scanning infinite objects. For example ICAP scanning video/audio streams or large objects being uploaded/downloaded using the FTP proxy."

Our initial recommendation would be to never deploy a "single disk" (8 disk 1 + XYZ GB disk 2) ISG virtual machine and set proxy policies to avoid never ending or huge files to be sent to CAS for ICAP feedback (article ID: 166921 - "Queued ICAP, Slowness or latency").

From investigating the storage data in the appliance config, you would see the below.

__________________________________________________________________________

 

Hardware Information

URL_Path /Diagnostics/Hardware/Info

Hardware Information

 

Model: SG-Enterprise

RAM: 81920 MB

RAM slot #0: 16384 MB, pn:Unspecified

RAM slot #1: 16384 MB, pn:Unspecified

RAM slot #2: 16384 MB, pn:Unspecified

RAM slot #3: 16384 MB, pn:Unspecified

RAM slot #4: 16384 MB, pn:Unspecified

Number of physical CPUs: 8

Number of cores: 8

CPU frequency: 2999 MHz

Storage: 3 drives

Disk in slot 1: 8 GB VMware   Virtual disk    , rev:2.0  serial: attributes:boot status:present

Disk in slot 2: 214 GB VMware   Virtual disk    , rev:2.0  serial: attributes:data status:present

Disk in slot 3: 214 GB VMware   Virtual disk    , rev:2.0  serial: attributes:data status:present

Network: 

Interface 0:0: Virtual network   running with virtual link  (MAC 00:50:56:b9:ad:53)

Interface 1:0: Virtual network   running with virtual link  (MAC 00:50:56:b9:bb:85)

Interface 2:0: Virtual network   running with virtual link  (MAC 00:50:56:b9:67:34)

Interface 3:0: Virtual network   with no link  (MAC 00:50:56:b9:0d:e0)

Accelerators: none

__________________________________________________________________________

Resolution

To fix this challenge, we recommend to resize disk 2 & disk 3 to 2 * 214 GB, or add more disk. For the resizing details, please refer to the snippet below. The resource doc. is attached, for your reference.