When my Seafile server goes offline - SeaDrive shows the red error thing… but…
When the Seafile server comes back online, the seadrive client continues to show the red error everywhere (ie two separate pcs).
This is Seadrive 3.0.17 on windows, Server Version: 11.0.13 on ubuntu VM
Only way I get the seadrive notification to come back to green is by going to show filesync errors.
Shouldn’t this indicator update itself automatically? Can’t see any reason for it to keep reminding me the server was offline recently…
I take the Server offline regularly to backup the server VM - it is just an annoyance more than anything. Maybe i’m doing the backups wrong but its what works for me…
Hi,
this issue has been fixed in SeaDrive 3.0.22
good to hear I will try it out
still seem to be having the same issue after upgrading the seadrive client to 3.0.22 - same on the two PCs i have it stalled on. Not the end of the world but would be nice to have it reflect the actual online state.
Hello @BigDaddio ,
When you encounter this issue, could you check what error is reported in seadrive.log? You can right-click the tray icon and select “Open logs folder” to locate and access the seadrive.log file.
Here is some extra info - 7am is when the backup script runs
The seafile.log
[05/28/26 04:54:27] cache size limit is 10000000000
[05/28/26 05:04:27] cache size limit is 10000000000
[05/28/26 05:14:27] cache size limit is 10000000000
[05/28/26 05:24:27] cache size limit is 10000000000
[05/28/26 05:34:27] cache size limit is 10000000000
[05/28/26 05:44:27] cache size limit is 10000000000
[05/28/26 05:54:27] cache size limit is 10000000000
[05/28/26 06:04:27] cache size limit is 10000000000
[05/28/26 06:14:27] cache size limit is 10000000000
[05/28/26 06:24:27] cache size limit is 10000000000
[05/28/26 06:34:27] cache size limit is 10000000000
[05/28/26 06:44:27] cache size limit is 10000000000
[05/28/26 06:54:27] cache size limit is 10000000000
[05/28/26 07:00:47] Bad response code for GET https://seafile.tct.co.nz/seafhttp/accessible-repos/?repo_id=0c51d5f1-c573-4fdb-a2ab-bc5c89809812: 502.
[05/28/26 07:00:47] Bad response code for GET https://seafile.tct.co.nz/seafhttp/accessible-repos/?repo_id=0316898e-6cef-4872-98aa-8948fb9f9ba2: 502.
[05/28/26 07:01:17] Bad response code for GET https://seafile.tct.co.nz/seafhttp/accessible-repos/?repo_id=0c51d5f1-c573-4fdb-a2ab-bc5c89809812: 403.
[05/28/26 07:01:17] Bad response code for GET https://seafile.tct.co.nz/seafhttp/accessible-repos/?repo_id=0316898e-6cef-4872-98aa-8948fb9f9ba2: 403.
[05/28/26 07:01:17] Bad response code for GET https://seafile.tct.co.nz/api2/repos/: 500.
[05/28/26 07:01:17] Bad response code for GET https://seafile.tct.co.nz/api2/repos/: 500.
[05/28/26 07:04:27] cache size limit is 10000000000
[05/28/26 07:14:27] cache size limit is 10000000000
[05/28/26 07:24:27] cache size limit is 10000000000
[05/28/26 07:34:27] cache size limit is 10000000000
[05/28/26 07:44:27] cache size limit is 10000000000
[05/28/26 07:54:27] cache size limit is 10000000000
[05/28/26 08:04:27] cache size limit is 10000000000
[05/28/26 08:14:27] cache size limit is 10000000000
[05/28/26 08:24:27] cache size limit is 10000000000
[05/28/26 08:34:27] cache size limit is 10000000000
[05/28/26 08:44:27] cache size limit is 10000000000
[05/28/26 08:54:27] cache size limit is 10000000000
[05/28/26 09:04:27] cache size limit is 10000000000
[05/28/26 09:14:27] cache size limit is 10000000000
[05/28/26 09:24:27] cache size limit is 10000000000
The gui log
[05/27/26 13:13:38] failed to read request from shell extension: The pipe has been ended.
[05/27/26 13:13:59] failed to read request from shell extension: The pipe has been ended.
[05/28/26 07:00:49] QSystemTrayIcon::showMessage: Wrong icon size (42x42), please add standard one: 43x43
[05/28/26 07:01:16] request failed for https://seafile.tct.co.nz/api2/auth/ping/:
Page unavailable
Page unavailable
<p>Sorry, but the requested page is unavailable due to a server hiccup.</p>
<p>Our engineers have been notified, so check back later.</p>
[05/28/26 07:01:16] request failed for https://seafile.tct.co.nz/api2/account/info/:
Page unavailable
Page unavailable
<p>Sorry, but the requested page is unavailable due to a server hiccup.</p>
<p>Our engineers have been notified, so check back later.</p>
[05/28/26 07:01:16] request failed for https://seafile.tct.co.nz/api2/auth/ping/:
Page unavailable
Page unavailable
<p>Sorry, but the requested page is unavailable due to a server hiccup.</p>
<p>Our engineers have been notified, so check back later.</p>
[05/28/26 07:01:16] request failed for https://seafile.tct.co.nz/api2/account/info/:
Page unavailable
Page unavailable
<p>Sorry, but the requested page is unavailable due to a server hiccup.</p>
<p>Our engineers have been notified, so check back later.</p>
[05/28/26 07:01:19] QSystemTrayIcon::showMessage: Wrong icon size (42x42), please add standard one: 43x43
[05/28/26 07:04:17] request failed for https://seafile.tct.co.nz/api2/auth/ping/:
Page unavailable
Page unavailable
<p>Sorry, but the requested page is unavailable due to a server hiccup.</p>
<p>Our engineers have been notified, so check back later.</p>
this is a section of the backup code I use
mkdir -p “$FILES_DEST/SeafileSQLDump”
backup_db () {
DBNAME=$1
OUTFILE=“$FILES_DEST/SeafileSQLDump/${DBNAME}_$DATE.sql”
log “Dumping $DBNAME…”
docker exec “$DB_CONTAINER” \
mysqldump -u"$MYSQL_USER" -p"$MYSQL_PASSWORD" \
–single-transaction \
–skip-lock-tables \
–quick \
–routines \
–events \
–hex-blob \
–default-character-set=utf8mb4 \
“$DBNAME” > “$OUTFILE”
}
backup_db ccnet_db
backup_db seafile_db
backup_db seahub_db
after the DB backup I stop docker containers - rsync the /opt folder, then restart the docker containers
log “Backup /opt”
rsync -aAX --info=progress2 \
–exclude=‘/opt/seafile-data/seafile’ \
/opt “$FILES_DEST/”
Hope that helps - not the end of the world but would be nice to have it sorted…
Hello @BigDaddio ,
This issue was likely caused by not completely stopping all services before performing the backup. For example, you may have stopped Seafile or the database, but not Seahub. In that case, when Seahub tries to retrieve the library list, Nginx can still successfully access Seahub, but Seahub cannot communicate with Seafile or the database, resulting in a 500 error. Since a 500 error is considered a server-side error rather than a network error, the client will display it as a sync error.
I recommend fully stopping both Seafile and Seahub before starting the backup. This way, when the client accesses Nginx, it will receive a 502 error instead, which will be treated as a network error by the client.
Also, if you restart the database, make sure it is fully up and running before starting Seafile and Seahub again. Otherwise, incoming requests during startup may fail because the server cannot access the database, which could again result in non-network errors.
not sure I mentioned it but this is running in docker,
will it be enough to do a “docker stop seafile” container ie the whole container rather than targeting individually the seafile/seahub services?
It does look like my backup script was stopping the seafile-mysql container and not the seafile container itself, so that might look like a server error rather than a network error as you suggest?
UPDATE: so I went ahead and added a docker seafile stop before the docker seafile-mysql stop, and similarly at the other end to restart the docker containers and it appears to be working correctly now.
Appreciate the help