From 03980de072058d3b5a5c39484e005aaef7cdd8b5 Mon Sep 17 00:00:00 2001
From: Mathieu Leplatre <mathieu@mozilla.com>
Date: Thu, 23 Jul 2026 13:53:23 +0200
Subject: [PATCH 1/2] git-reader: gitupdate: show size on init and add more
 logs about tags

---
 git-reader/run.sh | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/git-reader/run.sh b/git-reader/run.sh
index c7fb9f22..87c09b37 100644
--- a/git-reader/run.sh
+++ b/git-reader/run.sh
@@ -13,7 +13,14 @@ log_sizes() {
     git_size=$(du -sh "$repo_path/.git/objects" 2>/dev/null | cut -f1) || true
     lfs_size=$(du -sh "$repo_path/.git/lfs" 2>/dev/null | cut -f1) || true
     log "Size of $repo_path: ${total_size} (git objects=${git_size:-n/a}, LFS=${lfs_size:-n/a})"
-    du -sh "$repo_path}/attachments/*/*" | sort -hr | head -n 10
+
+    log "Number of tags per collection (top 10):"
+    git -C "$repo_path" tag | sed 's|/[0-9]*$||' | sort | uniq -c | sort -rn | head -n 10
+    echo "..."
+
+    log "Size of attachments folders (top 10):"
+    du -sh "$repo_path"/attachments/*/* 2>/dev/null | sort -hr | head -n 10
+    echo "..."
 }
 
 ORIGIN_NAME="origin"
@@ -63,6 +70,8 @@ cmd_gitupdate() {
         log "Setting latest symlink to A..."
         ln -sfn "$repo_path/A" "$repo_path/latest"
 
+        log_sizes "$repo_path/latest"
+
         # A was just cloned at origin HEAD and B is an exact copy, so there is
         # nothing to update yet. Return and let the next run pick up changes.
         LOG_PREFIX=""

From 417e6f47caea2512a9c77449fbecf14b837f6b72 Mon Sep 17 00:00:00 2001
From: Mathieu Leplatre <mathieu@mozilla.com>
Date: Thu, 23 Jul 2026 14:14:13 +0200
Subject: [PATCH 2/2] Show number of files too

---
 git-reader/run.sh | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/git-reader/run.sh b/git-reader/run.sh
index 87c09b37..92b32aa6 100644
--- a/git-reader/run.sh
+++ b/git-reader/run.sh
@@ -14,6 +14,11 @@ log_sizes() {
     lfs_size=$(du -sh "$repo_path/.git/lfs" 2>/dev/null | cut -f1) || true
     log "Size of $repo_path: ${total_size} (git objects=${git_size:-n/a}, LFS=${lfs_size:-n/a})"
 
+    local file_count lfs_count
+    file_count=$(git -C "$repo_path" ls-files | wc -l)
+    lfs_count=$(git -C "$repo_path" lfs ls-files | wc -l)
+    log "Number of files: ${file_count} (LFS=${lfs_count})"
+
     log "Number of tags per collection (top 10):"
     git -C "$repo_path" tag | sed 's|/[0-9]*$||' | sort | uniq -c | sort -rn | head -n 10
     echo "..."