• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Forwarded to devs Bug with Acronis Backup with Obsidian 18.0.40

besim

New Pleskian
Username:

TITLE

Bug with Acronis Backup with Obsidian 18.0.40

PRODUCT, VERSION, OPERATING SYSTEM, ARCHITECTURE

Plesk Obsidian v18.0.40_build1800211119.12 os_Debian 10.0, Intel(R) Xeon(R) Gold 5120 CPU @ 2.20GHz (4 core(s))

PROBLEM DESCRIPTION

The extension "Acronis Backup" (v1.5.5) is used, so far without problems. A nightly incremental backup is created to an external Acronis backup server.

After updating to Plesk Obsidian 18.0.40, an error is reported in the Acronis Dashboard after each backup, which is also sent via mail to the administrator.

STEPS TO REPRODUCE

After updating to Plesk Obsidian 18.0.40, an error is reported in the Acronis Dashboard after each backup, which also goes via mail to the administrator:

The 'Executing pre-data capture command' activity failed. The pre-data capture command completed with the following 'non-zero' return code: 1

In the detailed log entry in the Acronis console about this error, there is a more detailed description that reports incorrect execution of an internal command that is executed as part of the backup. This script is unedited and included with the client.

{ "code": "error", "error": { "domain": "environment", "code": "MemoPreDataCaptureCommandNonZeroReturnCode", "reason": "MemoPreDataCaptureCommandNonZeroReturnCode", "context": { "ExitCode": 1, "_src": { "code": 20250646, "fields": { "$module": "service_process_lxa64_27904", "CommandID": "2F224F57-6933-4BFC-A6E9-E9E5A47DBA2D" }, "src": { "file": "d:/35/enterprise/common/tol/command/command.cpp", "func": "MakeFailResult", "line": 495, "tag": "0x8d165e86fb8195bd" }, "suberror": { "code": 20250646, "fields": { "$module": "disk_bundle_lxa64_27904", "CommandID": "2F224F57-6933-4BFC-A6E9-E9E5A47DBA2D" }, "src": { "file": "d:/35/enterprise/common/tol/command/command.cpp", "func": "MakeFailResult", "line": 495, "tag": "0x8d165e86fb8195bd" }, "suberror": { "code": 1024, "fields": { "$module": "mms_lxa64_27904", "ExitCode": 1 }, "src": { "file": "d:/35/processor/spawn_operation.cpp", "func": "RunProcess", "line": 882, "tag": "0x858cc05f488a8e52" }, "suberror": { "code": 10944515, "fields": { "$module": "mms_lxa64_27904" }, "src": { "file": "d:/35/enterprise/mms/managers/cmdline/impl/manager.cpp", "func": "Execute", "line": 93, "tag": "0x6bf76712d83410aa" }, "suberror": {}, "text": "Failed to execute command '/usr/lib/Acronis/BackupAndRecovery/webcpprecapture: /usr/bin/sw-engine: /usr/lib/Acronis/system_libs/8.2.1/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /usr/bin/sw-engine)\n/usr/bin/sw-engine: /usr/lib/Acronis/system_libs/8.2.1/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /lib/x86_64-linux-gnu/libaps.so.1)\n/usr/bin/sw-engine: /usr/lib/Acronis/system_libs/8.2.1/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /usr/lib/x86_64-linux-gnu/libboost-plesk-1.74/libboost_regex-plesk.so.1.74.0)\n/usr/bin/sw-engine: /usr/lib/Acronis/system_libs/8.2.1/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /lib/x86_64-linux-gnu/liblock_manager.so.2)\n/usr/bin/sw-engine: /usr/lib/Acronis/system_libs/8.2.1/libstdc++.so.6: version `GLIBCXX_3.4.26' not found (required by /usr/lib/x86_64-linux-gnu/sw/librdbmspp.so.2)\nexit status 1\n'.",

ACTUAL RESULT

A backup is then created, but I cannot check whether this backup is complete. However, it is accessible as far as it goes. The above error messages are generated each time a backup is attempted (whether automatically or manually).

EXPECTED RESULT

None of these error messages and a correct backup.

ANY ADDITIONAL INFORMATION

none

YOUR EXPECTATIONS FROM PLESK SERVICE TEAM

Confirm bug
 
Addendum:

Within Plesk, access to backups created in Acronis Backup does NOT work. The following error message is displayed:

An error has occurred. Try again later or contact your administrator.

However, the backup can be accessed via Acronis Console.
 
In the changelog I saw that they have changed the default compressor from Gzip to ZSTD and a couple of other "improvements" for the backup.
Glad I never update any server automatically when it releases, instead I'm waiting two weeks and in that time I'm searching the web for bug reports....
 
This is a bug on Acronis side. Acronis backup agent runs Plesk utilities with the custom value of LD_LIBRARY_PATH environment variable, which causes incorrect dynamic linking of Plesk code to Acronis own version of libstdc++ library instead of Plesk's version. This problem was triggered by the update of Plesk's library in 18.0.40.

Note that the connection of Acronis Backup extension to Acronis cloud service is broken too due to the same reason.

Acronis KB article on this issue describes the workaround: Acronis Backup extension for Plesk: Backup fails after Plesk update to version 18.0.40 | Knowledge Base.
 
Are there any updates on this? I'm am having the same issues with Acronis not being able to backup since the 18.0.40 upgrade.
 
Are there any updates on this? I'm am having the same issues with Acronis not being able to backup since the 18.0.40 upgrade.

Did you apply the workaround mentioned in the Acronis KB article above?
The issue will be resolved once Acronis has released a new version of the Acronis agent
 
I did, however, nothing changed and I cannot get Acronis to restart on my server:

Red Hat Enterprise Linux Server 7.9 (Maipo) Plesk Obsidian Version 18.0.40
 
Just FYI: There are now two knowledge base articles at Acronis with this bug.
 
Just FYI: There are now two knowledge base articles at Acronis with this bug.
I now have several servers with the latest Plesk software version and not 1 wants to update anymore.
The method that Acronis indicates in their documentation also does not work.
But the worst part is that you don't read anywhere when they hope the new update or patch is out.
 
The update will not come with Plesk, but with the Acronis client. This is managed by the respective backup server. So if a corresponding update of the client is published, then the administrator of the backup server is responsible for deploying the client. Currently, however, there is no update that fixes the error.

The workarounds work for me though.
 
The update will not come with Plesk, but with the Acronis client. This is managed by the respective backup server. So if a corresponding update of the client is published, then the administrator of the backup server is responsible for deploying the client. Currently, however, there is no update that fixes the error.

The workarounds work for me though.
My mistake, I also meant that I expected Acronis to come up with an update for its agent.
I only write it on this forum for the Plesk users.
 
Thought I'd share here as we had Acronis start failing again today on a box and thought it was the same issue again but wasn't

We troubleshooted by running webcpprecapture command manually in console to see what was happening
Internal PHP plesk was running out of memory and the php process was crashing

We upgraded as per this https://support.plesk.com/hc/en-us/...f-X-bytes-exhausted-tried-to-allocate-Y-bytes-

Problem solved - the default was only set to 256mb which is not particularly high for a box that's quite busy so wouldn't surprise me if other weirdness we get from time to time is being caused by the low-ish memory limit.

This was the error is was generating when run manually that didn't show up in the Acronis log

[2022-01-06 15:18:47.932] 22363:61d67bad3e1f5 INFO [extension/acronis-backup] Collecting metadata. DB servers: OK
PHP Fatal error: Allowed memory size of 268435456 bytes exhausted (tried to allocate 84411856 bytes) in /opt/psa/admin/plib/modules/acronis-backup/library/MetaDataWriter.php on line 28
ERROR: Allowed memory size of 268435456 bytes exhausted (tried to allocate 84411856 bytes) (MetaDataWriter.php:28)
[2022-01-06 15:18:48.319] 22363:61d67bad3e1f5 DEBUG [panel] Execute: '/opt/psa/admin/bin/send-error-report'
 
For our records: The problem was fixed with an official update (version 610) of the Acronis CyberGuard Client.
 
Back
Top