05-21-2021 05:08 AM - edited 05-26-2021 05:45 AM
Hi all,
Backup session sometimes ends with Network connection broken (40) error
The parameter was applied: MAX_ENTRIES_PER_ADD (25000). The problem was not solved
The problem has begun to manifest itself significantly in the last 2 weeks.Netbackup Configurations not changed
How can you solve this problem, or significantly reduce its manifestation?
Logs in attachment
Andrii
05-21-2021 05:57 PM - edited 05-21-2021 05:58 PM
Couple of things to consider in order to isolate this issue. 1st how big are the data you are trying to backup? are there any other task been done on the server during backup window or while backup is running? is backup failing for entire drives or filesystems or affecting a specific drive/filesystem?
05-25-2021 12:57 AM
Backups can often run concurrently
The error can occur when executing various types of policies: windows, oracle, vmware. The amount of data is different. There is no pattern
Veritas explains this issue: https://www.veritas.com/support/en_US/article.100047973. But the error rate has increased in the last 2 weeks. I admit that delays, timeouts in the work of the master server can indirectly affect the appearance of the error
05-25-2021 08:51 AM
Has anything changed in your network path for this client backup?
Fire Wall changes recently? Network bandwidth bottleneck due to heavier data loads from multiple hosts? There's several things to check/research on this issue. If you've checked your timeout settings and they appear to be long enough for things like client read timeout and etc....then I'd start looking at the network path(s) to/from the client host.
05-26-2021 07:02 AM
I had a convenient opportunity to analyze the error. Media server and client are on the same network. Small data backup failed with error 40. In this case, as well as in similar cases, the media server terminates the connection with the master server.
master_server/bpdbm
07:19:02.022 [11748.1432] <2> daemon_proxy_proto: Preparing to do daemon protocol for (ip_address_master:1556 <- ip_address_media10:63939)
07:19:02.038 [11748.1432] <16> vnet_proxy_handoff_receive: WSASocket() failed: 10107
07:19:02.038 [11748.1432] <16> vnet_proxy_protocol_from_legacy: vnet_proxy_handoff_receive() failed: 9
07:19:02.038 [11748.1432] <16> vnet_proxy_socket_swap: vnet_proxy_protocol_from_legacy() failed: 9
07:19:02.038 [11748.1432] <16> daemon_proxy_proto: vnet_proxy_socket_swap() failed: vnet status 9, nb status 23
07:19:02.038 [11748.1432] <16> process_request: Secure proxy protocol failed
media_server10/bpbrm:
07:19:02.032 [6304.2176] <2> ts_put_length_string_optimized: put n bytes failed: An established connection was aborted by the software in your host machine. (10053)
07:19:02.032 [6304.2176] <2> db_senddata: ts_put_string_handle(): connection dropped or not connected, An established connection was aborted by the software in your host machine. , (10053)
07:19:02.032 [6304.2176] <2> db_startrequest: db_sendrequest() failed: network connection broken
07:19:02.032 [6304.2176] <16> db_begin: db_startrequest() failed: network connection broken
07:19:02.032 [6304.2176] <2> db_FLISTsend: db_begin() failed: network connection broken
07:19:02.048 [6304.2176] <16> non_mpx_backup_archive_verify_import: db_FLISTsend failed: network connection broken (40)
05-27-2021 03:13 AM
What version do you use? Veritas has TechNote https://www.veritas.com/support/en_US/article.100048099 and mentions EEBs for the issue,
05-27-2021 05:10 AM
Version 8.1.1 is used
Thanks for the link. We have tried to follow these guidelines
EEB was not used.