From Newsgroup: news.software.readers

Hello everyone,

Please be advised that we will be performing critical maintenance on our news server tonight.

The service will be shut down at midnight (00:00 CEST, Paris/DE time, 22:00 UTC) on September 23rd. The downtime is necessary to repair a corruption in the overview database (`tradindexed`) which is currently blocking the propagation of all new articles.

We expect the maintenance to last for several hours. The NNTP service will be unavailable during this period.

We apologize for the inconvenience and thank you for your understanding.

Gabx

--- Digital Signature --- TAPJXTTBJsKX8n3ml02a3V4afj3vos9TTA5QvYNftv1y9RXcVVHXQ/1i5zPBxcjVdpcQ2sHdVDTVsVU3+8qkBQ==


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: news.software.readers

In article <20250922.1758554012.435992.4088@m2usenet.local>
Gabx <administrator@news.tcpreset> wrote:

Hello everyone,

Please be advised that we will be performing critical maintenance on our news server tonight.

The service will be shut down at midnight (00:00 CEST, Paris/DE time, 22:00 UTC) on September 23rd. The downtime is necessary to repair a corruption in the overview database (`tradindexed`) which is currently blocking the propagation of all new articles.

We expect the maintenance to last for several hours. The NNTP service will be unavailable during this period.

We apologize for the inconvenience and thank you for your understanding.

Gabx

No worries, and thanks for what you do.

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: news.software.readers

Gabx wrote:

Hello everyone,

The propagation failure was caused by a corrupt overview database
(tradindex), which prevented the server from accepting any new articles.

During the downtime, we performed a full rebuild of the overview
database directly from the article spool.

Server logs look stable, and my initial monitoring shows that the
underlying corruption is resolved and reader clients can access articles correctly.

We are now monitoring the news flow and would appreciate it if users and
peers could confirm if they see propagation from our site returning to
normal. Your feedback would be very helpful in confirming that the issue
is fully resolved for everyone.

Thank you for your patience during this process.


Gabx

https://news.tcpreset.net
--
0745 074D FEAA 9CB7 62E9 D89D 3E54 F490 F2CC 5A82
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: news.software.readers

On 23.09.25 02:33, Gabx wrote:

Gabx wrote:

Hello everyone,

The propagation failure was caused by a corrupt overview database (tradindex), which prevented the server from accepting any new articles.

I crashed tradindex and tradspool more than once
just by doing nothing...I don't know what...

My preferred overview method is ovsqlite and cyc buffs storage.

Using ZFS w/ compression=lz4 can do near 3x and cost no performance
and did not break yet :D
--
.......
Billy G. (go-while)
https://pugleaf.net
@Newsgroup: rocksolid.nodes.help
irc.pugleaf.net:6697 (SSL) #lounge
TOR-IRC: ij7fmstcudrwyty5p6iloodeenur4wksovlhbyrfubat7eixplmrxvqd.onion:6667 --- Synchronet 3.21a-Linux NewsLink 1.2